Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "Interpersonal diagnosis of personality; a functional theory and methodology for personality evaluation"

See other formats





Timothy Leary 

Interpersonal Diagnosis 



A Functional Theory and Methodology 
for Personality Evaluation 





Resource Publications 

An imprint of Wipf and Stock Publishers 
199 West 8th Avenue • Eugene OR 97401 

Resource Publications 

A division of Wipf and Stock Publishers 

199W8th Ave, Suites 

Eugene, OR 97401 

Interpersonal Diagnosis of Personality 

A Functional Theory and Methodology for Personality Evaluation 

By Leary, Timothy 

Copyright© 195 7 by Leary, Timothy 

ISBN: 1-59244-776-7 

Publication date 7/30/2004 

Previously published by John Wiley & Sons, 1957 

Marianne Leary 


This book is concerned with interpersonal behavior, primarily as ex- 
pressed and observed in the psychotherapeutic setting. Its value lies 
in its emphasis on the complexity and variety of human nature and 
on the objectivity and clarity of the empirical procedures it sets forth 
for multilevel diagnosis. The research on which it reports was made 
possible by grants from the United States Public Health Service and 
the Kaiser Foundation. 

The interpersonal factors of personality are those conscious or 
unconscious processes which people use to deal with others and to 
assess others and themselves in relation to others. The aim of the inter- 
personal machinery of personality is to ward off anxiety and preserve 
self-esteem. One of the major results of these operations is to create 
the social environment in which each person lives. 

Everyone tends to make his own interpersonal world. Neurosis 
or maladjustment involves the limiting of one's interpersonal appara- 
tus and the compulsive use of certain inflexible, inappropriate inter- 
personal operations which bring about results that are painful, unsatis- 
factory, or different from one's conscious goals. Adjustment is char- 
acterized by an understanding of one's personahty structure, by the 
development of mechanisms flexible enough to deal with a variety of 
environmental pressures, and by the management of one's behavioral 
equipment in such a way as to avoid situations where the mechanisms 
will be ineffective or damaged. 

Any statement about human nature, however, is restricted in 
meaning unless the level of behavior to which it refers is made clear. 
The first step must be a definition of levels and an ordering of data in 
terms of levels. The aim of the research work described in this book 
has been to develop a multilevel model of personality and to present a 
series of complex techniques for measuring interpersonal expressions 
at these different levels of personality. A conceptual and empirical 
method for converting observations of interpersonal behavior is set 
forth. The reader will encounter new theories about the effect of 
interpersonal behavior, the meaning of fantasy expressions, the social 
language of symptoms, and the nature and functional meaning of con- 
flict. These theories and systematic procedures constitute the Inter- 


personal System of Peisonality, developed by the Kaiser Foundation 
Psychology Research Project. 

The approach employed might be called a dynarmc behaviorism. 
There are two dynamic attributes. The first refers to the impact one 
person has or makes in interaction with others; the second refers to 
the interaction of psychological pressures among the different levels 
of personality. The behavioristic attributes of the system derive from 
the procedure of viewing every response of the subject (overt, verbal, 
symbolic) as a unit of behavior which is classified by objective 
methods and automatically sorted into the appropriate level of per- 
sonality. The patterns and clusters of thousands of these responses, 
sorted into different levels, are then converted by mathematical tech- 
niques into indices and into a multilevel diagnostic code summary. 
These are then related to clinical events or prognoses. In the develop- 
ment of the interpersonal system more than 5,000 cases (psychiatric, 
medical, and normal controls) have been studied and diagnosed. 

In addition to describing and validating the process of interper- 
sonal diagnosis in the psychiatric clinic, this volume demonstrates how 
these theories and methods may be applied in four other practical set- 
tings — in the psychiatric hospital, in psychosomatic medicine, in 
industrial management, and in group therapy. 

This book should be interpreted in the light of its environmental 
and professional contexts. It is the product of clinical psychologists 
working in a psychiatric setting, and practical answers have been 
required of the interpersonal system at each stage of its development. 
This gives the book its functional cast. As to its implications for the 
profession of psychology, in my own mind at least, a new concept of 
the "clinical psychologist-as-diagnostician" has emerged. In the 
Introduction, I have detailed the genesis of the research which has 
resulted in the book, and have set forth the contributions of the many 
people who have helped to bring it to fruition. 

Timothy Leary 

Berkeley, California 
October, 1956 



Part I 

Some Basic Assumptions About Personality 


1. Interpersonal Dimension OF Personality 3 

2. Adjustment-iMaladjustment Factors in Personality Theory 17 

3. Systematizing the Complexity of Personality . . .33 

4. Empirical Principles in Personality Research ... 45 

5. Functional Theory of Personality 50 

6. General Survey of the Interpersonal and 

Variability Systems 59 

Part II 

The Interpersonal Dimension of Personality: 
Variables, Levels, and Diagnostic Categories 

Introduction 90 

7. The Level of Public Communication: 

The Interpersonal Reflex 91 

8. The Level of Conscious Communication: 

The Interpersonal Trait 132 

9. The Level of Private Perception: 

The Interpersonal Symbol 154 

10. The Level of THE Unexpressed: Significant Omissions . . 192 

11. The Level of Values: The Ego Ideal 200 

12. A System of Interpersonal Diagnosis 207 


Part III 

The Variability Dimension of Personality: 
Theory and Variables 

Introduction 240 

13. The Indices OF Variability .241 

Part IV 
Interpersonal Diagnosis of Personality 

14. Theory of Multilevel Diagnosis 265 

15. Adjustment Through Rebellion: The Distrustful 

Personality 269 

16. Adjustment Through Self-Effacement: 

The Masochistic Personality 282 

17. Adjustment Through Docility: The Dependent 

Personality 292 

18. Adjustment Through Cooperation: 

The Overconventional Personality 303 

19. Adjustment Through Responsibility: 

The Hypernormal Personality 315 

20. Adjustment Through Power: The Autocratic 

Personality 323 

21. Adjustment Through Competition: The Narcissistic 

Personality 332 

22. Adjustment Through Aggression: The Sadistic Personality 341 

Part V 
Some Applications of the Interpersonal System 

Introduction 352 

23. Interpersonal Diagnosis of Hospitalized Psychotics . .354 

24. Interpersonal Diagnosis in Medical Practice: 

Psychosomatic Personality Types 373 

25. Analysis of Group Dynamics in an Industrial 

Management Group 403 

26. Predicting and Measuring Interpersonal Dynamics in 

Group Psychotherapy 426 




1. Illustrations of the Measurement of Interpersonal 

Behavior at Level I 439 

2. The Interpersonal Adjective Check List . . . .455 

3. The Administration, Scoring, and Validation of the 

Level III-TAT 464 

4. The Interpersonal Diagnostic Report 480 

5. Norms, Conversion Tables, and Weighted Scores Used in 

Interpersonal Diagnosis 493 

Index of Names 501 

Index of Subjects 503 


1. Continuum of the Sixteen Interpersonal V^ariables .... 65 

2. Interpersonal Behavior in Psychotherapy 68 

3. Change in Behavior in Therapy 70 

4. Seven Generic Areas of Personality 85 

5. Interpersonal Interactions in Group Therapy 95 

6. Categorization of Check-List Items 135 

7. Illustrative Self-Diagnosis 138 

8. Conscious Description of Father 139 

9. Conscious Description of Mother 140 

10. Conscious Description of Self and Family Members .... 141 

11. Patient's Description of Therapist 142 

12. Therapist's Description of Patient 144 

13. Pattern of Familial Relations 146-47 

14. Variation in Depth of Measures 151 

15. Diagnosis of Walter Mitty 174 

16. Profile of TAT Scores 176 

17. Conscious and "Preconscious" Profiles 178 

18. Conflict Between Power and Weakness 179 

19. Facade of Power and Responsibility 180 

20. Facade of Weakness and Docility 180 

21. Rigidly Conventional Profiles 183 

22. Depth Continuum of Personality Levels 187 

23. Rigid Avoidance of Rebelliousness 194 

24. Consistent Omission of Rebellious Themes 196 

25. Docile Subject Idealizes Strength 204 

16. Summary Scores for Overconventional Patient 218 

27. Diagnosis of Facade Behavior 219 

28. Illustration of Interpersonal Diagnosis 222-23 

29. Diagnosis of Level III Behavior 224 

30. Illustration of Multilevel Diagnosis 226-27 

31. Generic Variability Indices 250 

32. Calculation of Discrepancy Values 258 

33. Behavior of Ten Samples at Level I 380 




34. Behavior of Ten Samples at Level II 382 

35. Behavior of Ten Samples at Level III 384 

36. Multilevel Mean Scores of Normal Controls 387 

37. Multilevel Mean Scores of Ulcer Patients 388 

38. Multilevel Mean Scores of Hypertensive Patients 390 

39. Multilevel Mean Scores of Obese Women 392 

40. Overtly Neurotic Dermatitis Patients 394 

41. Self-inflicted Dermatitis Patients 395 

42. Unanxious Dermatitis Patients 397 

43. Psychiatric Clinic Sample 398 

44. Multilevel Mean Scores of "Neurotics" 399 

45. Multilevel Mean Scores of "Psychotics" 401 

46. Self-Descriptions of Four Executives 406 

47. Self-Deception Indices of Four Executives 407 

48. Group Dynamics Booklet 411-17 

49. Perceptions by General Manager 418 

50. Consensual Diagnosis of General Manager 420 

51. Perceptions by Production Manager 421 

52. Consensual Diagnosis of Production Manager 422 

53. Perceptions by Personnel Manager 423 

54. Consensual Diagnosis of Personnel Manager 424 

55. Predictions of Interpersonal Roles 429 

56. Measurements of Interpersonal Roles 430 

57. Diagram of Five Measures of Personality 432-33 

58. Two Contrasting MMPI Profiles 442 

59. The Level I Diagnosis 444 

60. Summaries of Interpersonal Behavior 452 

61. The Diagnostic Booklet 482-88 

62. Multilevel Profile Before and After Psychotherapy . . . .491 



1. Operational Definition of Five Levels of Personality . . . .81 

2. Percentage of Diagnostic Types (Level I-M) 129 

3. Percentage of Diagnostic Types (Level II-C) 152 

4. Illustrative Classification of Interpersonal Behavior at the Symbolic 

or Projective Level 170 

5. Percentage of Diagnostic Types (Level III-T) 190 

6. Three Elements of Diagnosis of Personality: Classification, 

Profiles, and Report 214 

7. The Adaptive and Maladaptive Interpersonal Diagnostic Types . . 220 

8. Median Interpersonal Self-Description Score for Six MMPI 

Clinical Groups 231 

9. Operational Redefinition of Psychiatric Categories in Terms of 

Interpersonal Operations 233 

10. Informal Listing of the Twelve Generic Variability Indices . . .252 

11. Operational Definition of Forty-eight Indices of Variation . . 254-56 

12. Key to Numbers and Letters Employed in Coding Variability Indices 256 

13. Horizontal (Lov) and Vertical (Dom) Values for Each Octant . . 260 

14. All Possible Discrepancies Around the Pair \-l and Their Magnitudes 260 

15. Illustration of the Grouping of All Possible Discrepancies Involving 

the Diagnostic Codes 1 and / 261 

16. Percentage of Rebellious-Distrustful Personalities (Level I-M) . . 280 

17. Percentage of Rebellious-Distrustful Personalities (Level II-C) . .281 

18. Percentage of Self-Effacing-Masochistic Personalities (Level I-M) . 290 

19. Percentage of Self-Effacing-Masochistic Personalities (Level II-C) . 291 

20. Percentage of Docile-Dependent Personalities (Level I-M) . . . 299 

21. Percentage of Docile-Dependent Personalities (Level II-C) . . 300 

22. Percentage of Cooperative-Overconventional Personalities 

(Level I-M) 312 

23. Percentage of Cooperative-Overconventional Personalities 

(Level II-C) 313 

24. Percentage of Responsible-Hypernormal Personalities (Level I-M) . 321 

25. Percentage of Responsible-Hypernormal Personalities (Level II-C) . 322 

26. Percentage of Managerial-Autocratic Personalities (Level I-M) . .330 



TABLE ^^^^ 

27. Percentage of Managerial- Autocratic Personalities (Level II-C) . 331 

28. Percentage of Competitive-Narcissistic Personalities (Level I-M) . 338 

29. Percentage of Competitive-Narcissistic Personalities (Level II-C) . 340 

30. Percentage of Aggressive-Sadistic Personalities (Level I-M) . . 349 

31. Percentage of Aggressive-Sadistic Personalities (Level II-C) . .350 

32. Level I Diagnoses Assigned to 148 Patients in the Three 

Psychotic Samples 356 

33. Level II-C Diagnosis of 46 Patients in the Three Psychotic Samples . 357 

34. Level III-T Diagnosis of 38 Patients in the Three Psychotic Samples 358 

35. The Significance of Differences Among Ten Symptomatic Groups 

at Level I-M 381 

36. The Significance of Differences Among Ten Symptomatic Groups 

at the Level of Conscious Self-Description (Level II-C) . . 383 

37. The Significance of Differences Among Ten Symptomatic Groups 

at the Level of "Preconscious" Expression (Level Ill-T [Hero]) . 385 

38. Illustrative Calculation of MMPI Indices for Measuring Symptomatic 

Behavior (Level I-M) 443 

39. Illustration of the Calculations for Determining the Level I Profile for 

a "Neurotic" Patient, SN 453 

40. Interpersonal Check List, Form 4, Words Arranged by Octant and 

Intensity 456-57 

41. Test-Retest Correlations, Form IIIa, by Octant and Sixteenth . . 461 

42. Average Intervariable Correlation as a Function of Their Separation 

Around the Circle 462 

43. ICL Means and Standard Deviations for Psychiatric Outpatients . . 463 

44. Guide to Assigning Interpersonal Ratings to Ten TAT Stories 

(Level III-T) 466 

45. Molar Rating Sheet 471 

46. Means and Sigmas of Normative Group for Level III-T Hero and 

"Other" 472 

47. Chi-Square Relating the Kind of Initial Discrepancy on Dominance- 

Submission Between Conscious Self-Diagnosis and TAT Diagnosis 
to the Kind of Change in Self-Diagnosis of Dominance-Submission 
on Pre-Post Tests for 23 Psychotherapy Patients .... 474 

48. Chi-Square ... for 40 Discussion Group Controls .... 475 

49. Chi-Square ... for Combined Samples of 23 Psychotherapy Patients 

and 40 Obesity Patients 475 

50. Chi-Square Relating the Kind of Initial Discrepancy on Love-Hostility 

Between Conscious Self-Diagnosis and TAT Diagnosis to the Kind 
of Change in Self-Diagnosis of Love-Hostility on Pre-Post Tests 
for 23 Psychotherapy Patients 476 




51. Chi-Square ... for 40 Discussion Group Controls .... 476 

52. Chi-Square ... for Combined Samples of 23 Psychotherapy Patients 

and 40 Obesity Patients 477 

53. Chi-Square Relating the Amount of Discrepancy Between Conscious 

Self-Diagnosis and TAT Diagnosis to Amount of Temporal Change 

in Self-Diagnosis for 81 Discussion Group Controls .... 478 

54. Norms for Converting Raw Scores (Dom and Lov) to Standard 

Scores at Level I-M 494 

55. Norms for Converting Raw Scores (Dom and Lov) to Standard 

Scores at Level II-C 495 

56. Norms for Converting Raw Scores (Dom and Lov) to Standard 

Scores at Level III-TAT (Hero) 496 

57. Norms for Converting Raw Scores (Dom and Lov) to Standard 

Scores at Level III-TAT (Other) 497 

58. Weighted Scores for Measuring Discrepancy Between Two Diag- 

nostic Codes Indicating Kind and Amount of Difference Between 
Levels or Tests, and for Comparing Codes Where One Is of 
Extreme and the Other of Moderate Intensity .... 498-99 


In the past, the complexity of personality data, particularly as it is 
observed in the clinical setting, has led to a relative neglect of em- 
pirical studies and to an emphasis on anecdotal, speculative accounts. 
Where objective investigations have been undertaken, they have 
tended to be analyses which employed a single testing instrument. 
This is a result of the sociological development of the testing psy- 
chologist's role. 

The original and basic aim of the Kaiser Foundation Psychology 
Research was (and still is) the study of "process in psychotherapy." 
The first steps in this direction involved the construction of a sys- 
tematic way of viewing personality structure before therapy. This 
model system is necessary to predict what will happen in therapy and 
to measure change in structure during and after therapy. This book 
presents such a system and some of its diagnostic and prognostic 

The United States Public Health Service supported the research 
project by a series of six annual grants, from 1950 to 1954, under the 
directorship of Hubert S. Coffey and Dr. Saxton T. Pope, Jr., and 
from 1954 to 1956 under the direction of Timothy Leary. In addition 
to serving as the first director, Hubert Coffey has been chief advisor 
since the first days of the project. Dr. Pope provided research 
facilities and clinical wisdom, and was of signal help in developing the 
concept of variability indices, discussed in Chapter 13. 

The Kaiser Foundation contributed substantially to the research 
during the years 1950-1954, and from November 1954 assumed major 
support of the core project. Dr. Harvey Powelson became the 
director of the research project in 1951. He has given clinical advice, 
theoretical counsel, and administrative support throughout the dura- 
tion of the research. 

In its development, the interpersonal system of personality has 
been influenced by many collaborating psychologists and psychiatrists. 
It is impossible, in a cooperative, creative enterprise of this scope to 
accord specific credit for all contributions, and the following acknowl- 
edgments indicate only the major indebtedness. Those whose names 
are listed below should not, however, be held accountable for any 


weaknesses in the theoretical design. Full responsibility for the present 
version of the system is assumed by the author. 

The basic notion of the interpersonal classification system (the 
circle) was developed in 1948-1949 by Hubert Coffey, Mervin Freed- 
man, Timothy Leary, and Abel Ossorio. The same group was respon- 
sible for the original tripartite definition of levels. The psychotherapy 
groups which provided the original data for classification of inter- 
personal reflexes were organized with the help and sponsorship of 
J. Raymond Cope, of the Unitarian Church of Berkeley. 

Dr. Mary Sarvis, Kaiser Foundation Psychiatric Clinic, lent her 
diagnostic and therapeutic knowledge to the research group with un- 
sparing generosity. 

Mervin Freedman was a major participant in every stage of theo- 
retical and methodological development from 1948 to 1953. His 
thoughtful, analytic approach provided balance and good sense. 

Rolfe LaForge is responsible for the successful aspects of the 
statistical and methodological work. From 1950 to 1954, he directed 
the testing program, the IBM research, the check-list studies, and 
served as statistical consultant. 

Martin Levine, Blanche Sweet, Herbert Naboisek, and Ellen 
Philipsborn Tessman made theoretical contributions and aided in the 
processing of data. 

Jean Walker McFarlane was an original sponsor and advisor of the 
research project and contributed continuous editorial and practical 

Arthur Kobler of the Pinel Foundation Hospital, Seattle, has em- 
ployed the diagnostic system in his studies of psychotic patients. The 
combination of his empirical help and theoretical counsel has 
strengthened this book in several areas. 

Bernard Apfelbaum collaborated in the early stages of the oscilla- 
tion-variability theory. He also provided ratings of interpersonal 
behavior, as did Wanda Bronson, Albert Shapiro, and Marvin 

Frank Barron has served since 1950 as official and unofficial con- 
sultant to the research project. He helped design the original test 
battery and provided valuable editorial and methodological assistance. 
Psychotherapy groups studied by the research project were in 
charge of Dr. Jean Neighbor, Mary Darby Rauch, Shirley Hecht, 
Mervin Freedman, Stephen Rauch, Abel Ossorio, Dr. Harvey Powel- 
son, Robert Suczek, Hubert Coffey, Patrick SuUivan, and Richard V. 
Wolton. Richard Wolton also lent his assistance in the collection of 
data and in manuscript preparation. 


A most important aspect of the interpersonal system is that the 
test administration, scoring, and rating of tests — as well as the deter- 
mination of the multilevel diagnoses and the indices of conflict — are 
accomplished by highly trained technicians who are not professional 
psychologists. The technical staff responsible for the multilevel 
diagnoses of the 5,000 cases on which this book is based, includes Anne 
Apfelbaum, Elizabeth Asher, Mary della-Cioppa, Roberta Held, 
Charlotte Kaufmann, Joan Harvey LaForge, Helen Lane, and Bar- 
bara Lennon NichoUs. Gloria Best Martin was Research Administra- 
tor for the years 1950-1952. 

The countless administrative decisions necessary to maintain the 
day-to-day operations of the research project have been handled with 
competence by Miss Helen Lane. She has had final executive respon- 
sibility for data collection, office management, and manuscript 


Some Basic Assumptions About 
Personality Theory 

Interpersonal Dimension of Personality 

The twentieth century may well find historical status as the epoch in 
which man began to study himself as a scientific phenomenon. This 
development, inaugurated mainly by Sigmund Freud around the year 
1900, has brought about an impressive growth in the so-called human- 
ist disciplines — psychiatry, psychology, anthropology, sociology. The 
hour is yet too early to begin writing the chronicles of our time, but 
certain trends, now clearly evident, allow tentative predictions. 

The study of human nature appears, at this mid-century point, to 
be shifting from an emphasis on the individual to an emphasis on the 
individual-in-relation-to-others. During the last fifty years the sub- 
ject matter of psychiatry, for example, has moved away from case 
history and symptomatic labels and proceeded in the direction of 
social interaction analysis and psychocukural phenomena. The physi- 
calistic therapies, such as electro-shock and neurosurgery, seem to 
have worked with little theoretical justification against these scientific 
currents of the time.^ 

As late as twenty years ago the psychiatric literature was saturated 
with concepts that were oriented towards the nonsocial aspects of per- 
sonality — man in relation to his instinctual past (Freud), his racial past 
(Jung). The psychological laboratories at the same time buzzed with 
experiments on achievement, intelligence, temperament, and learning 
processes of the individual animal or human being. 

Today, theoretical events have taken a different turn. Man is 
viewed as a uniquely social being, always involved in crucial inter- 
actions with his family members, his contemporaries, his predecessors, 
and his society. All these factors are seen as influencing and being in- 
fluenced by the individual. The new direction is marked by a series of 
new conceptual guide posts from communication theory, cultural 
anthropology, and neop<;ychoanalysis. We possess a new bibliography 

* The research on neuropsychological relations accomplished at Tulane University 
under the direction of Robert G. Heath is a notable exception to this generalization. 


of guide books pointing out the approaching scientific horizons and 
relating them to the past. 

There is one concept which finds such wide and repeated expression 
in the current literature that it has taken on the debatable character 
of a motto. This is the term interpersonal relations. Introduced by the 
American psychiatrist Harry Stack Sullivan, it has become so popular 
that, at times, it appears destined to join those ill-fated concepts 
rendered meaningless by the frequency and pious generality of their 

The interpersonal theory is clearly a product of the converging 
theoretical trends of the time. It has many important implications for 
all the humanistic disciplines. 

This book and the research which it summarizes take as a starting 
point the interpersonal dimension of personality. We shall trace in 
the following chapters a theory, a measurement methodology, and a 
psychological diagnostic system based primarily on interpersonal be- 
havior. It seems appropriate, therefore, to take as the first question for 
consideration the definition of the basic term, interpersonal. 

What Is Interpersonal Behavior? 

Behavior which is related overtly, consciously, ethically, or sym- 
bolically to another human being (real, collective, or imagined) is 
interpersonal. This is a short but complex definition. Most of the 
succeeding pages will be devoted to its elaboration. 

Let us consider some examples of human behavior in the light of 
this definition. The report from a reliable observer "George insulted 
his father" is clearly interpersonal. It tells how George related to his 
father, what he did to his father. The finding "George says he is a 
friendly person" comes from a different observation point, the sub- 
ject's self-description, but is still clearly interpersonal. It tells how 
George perceives his motives toward other people. Also interpersonal 
is the inference made on the basis of dream or fantasy material "George 
dreams that his mother is protecting him." This refers to a fantasied 
relationship between the subject and another person. These descrip- 
tions of different aspects of the subject's behavior, which we call 
protocol statements, are the basic data on which we build a science of 
personality. They describe, at three different levels of observation, 
the subject's interpersonal relations. 

Another dimension of personality is reflected in the statements 
"George acts impulsively," "George says he is not depressed," "George 
dreams of hatboxes." These descriptions are taken from the same 
three levels of observation — the outsider's report, self-report, and 
dreams — but they are not directly interpersonal. Impulsivity, opti- 


mism, and a symbolic concern with containers have figured in certain 
personality theories and have some importance in the understanding 
of personality. Such descriptions are noninterpersonal because they 
do not refer to the subject's relationship to other people. They may 
be, and probably are, indirectly interpersonal. If we investigate 
further we might learn that George acts impulsively to impress others 
with his strength, that he says he is not depressed to prove that he 
does not need psychotherapy, and that he has a vague childhood 
memory of his mother bringing him lunch in a hatbox. The non- 
interpersonal thus becomes interpersonal; the personal characteristics 
take on a social meaning and reflect his relationships with others. 

We shall subsequently see that much of the conceptualization in 
psychology and the nomenclature of psychiatry has been noninterper- 
sonal. Terms such as depressed, impulsive, and inhibited, for example, 
refer to characteristics that possess maximum meaning when their inter- 
personal purpose is added. From the restricted and partisan inter- 
personal point of view, the functional value of such a popular diag- 
nostic phrase as "the patient acts depressed" is really not very great 
until we add, overtly or implicitly, the social implication. We make 
such a phrase more meaningful when we designate the interpersonal 
context or the interpersonal impact of the action — "to get the psy- 
chiatrist's sympathy" or "to show his parents how badly he feels 
they have treated him." 

Psychologists or psychiatrists who employ interpersonal concepts 
are generally characterized by an obsessive attention to the social im- 
plications of the subject's performance. They tend to view themselves 
as engaged in a complex relationship with the subject (or patient) and 
are particularly concerned with the social pressure which the subject 
is generating — the impression he is attempting to make upon them. 

The interpersonal psychologist generally carries away from an 
interview or a testing session a diagnosis centering not on the patient's 
intelligence or his symptoms, but rather on the social machinery which 
the patient put into action during the session. In most clinical situa- 
tions a numerical IQ index is of limited functional value. The clinician 
working from the interpersonal viewpoint would be more likely to 
stress not the patient's IQ, but the fact that "the patient acts in a wise 
manner and attempts to create the impression of intelligence," or, in 
another case, "the patient presents a fa9ade of docile simplicity, acts 
as though he were uninformed and eager-to-be-taught." 

SoTne Noninterpersonal Systems of Psychology 

The interpersonal system presented in this book addresses itself to a 
narrow, limited slice of human behavior. There are many other facets 


of human activity which have attracted the interest and energy of 
psychologists. In the Kaiser Foundation research we omit or ignore 
about nine-tenths of these activities and concentrate rather single- 
mindedly on one dimension — the interpersonal. We have restricted 
our theory to social behavior because we believe this to be the area 
of psychology which is most crucial and functionally important to 
human happiness and human survival. Our reasons for making this 
assertion will be detailed in a later section. 

In restricting our studies to one source of data we fail to take into 
account hundreds of important variables which characterize the 
individual. Height, weight, age, appearance, and motoric patterns are 
all factors which have some value in predicting behavior. All the 
physiological aspects of the individual are left out of our system. 

Sociological factors also contribute to the understanding of per- 
sonality and carry clear-cut interpersonal implications. We have been 
unable, so far, to include these factors in our investigations. 

Moreover, we have found it necessary to omit most of the variables 
which have had the highest priority for most psychologists — intelli- 
gence, interest patterns, political and culmral attitudes, and the 
variables of sensation and perception. 

Academic and experimental psychology has traditionally focused 
on the noninterpersonal aspects of behavior. Psychophysical experi- 
ments, learning theories, and intelligence and aptitude studies have 
monopolized the majority of the chapters in psychological texts. 
These areas are left completely untreated in the system of personality 
presented in this book. 

We are concerned, therefore, with a limited sector of the wide 
circle of human behavior. We concentrate simply on the way in 
which the individual deals with others — his actions, thoughts, fantasies, 
and values as they relate to others. In addition to restricting our atten- 
tion to interpersonal activity, there is a further qualification. We can- 
not hope to include the entire range of the individual's social behavior, 
but will apply most of our energies to the task of understanding and 
predicting the subject's interpersonal behavior in one specific environ- 
mental context — his relationship to a psychiatric clinic. 

Some Interpersonal Theories of Personality 

We have seen that in the last twenty years the cultural and social 
factors of human nature have become the object of widespread scien- 
tific attention. Sociologists and anthropologists have been actively 
applying psychiatric concepts to their data with mixed results. Entire 
primitive societies have been diagnosed as paranoid, or typed in terms 
of the ways in which they feed their young. 


At the same time, on the other side of the professional fence, several 
psychiatrists have assimilated the cultural into their thinking. Major 
revisions of orthodox Freudian concepts have developed. Three of 
the most successful of these personaHty theorists, Horney, Fromm, 
and Sullivan, have rejected the instinct theory and developed socially 
oriented structures of their own. A fourth, Erik H. Erikson, has con- 
structed an impressive system integrating social phenomena into the 
Freudian libido theory. 

Karen Horney began publishing in 1937 a series of important books 
in which she has developed a characterological approach to person- 
ality. She has described her dissatisfaction with the instinct theory and 
her own conceptual solutions in great detail. In her earliest work she 
contended that "neuroses are brought about by cultural factors" — 
which, more specifically, meant that neuroses are generated by dis- 
turbances in human relationships. 

In the years before I wrote The Neurotic Personality I pursued another line 
of research that followed logically from the earlier hypothesis. It revolved 
around the question as to what the driving forces are in neuroses. Freud had 
been the first to point out that these were compulsive drives. He regarded 
these drives as instinctual in nature, aimed at satisfaction and intolerance of 
frustration. Consequently he believed that they were not confined to neuroses 
per se but operated in all human beings. If, however, neuroses were an out- 
growth of disturbed human relationships, this postulation could not possibly be 
valid. The concepts I arrived at on this score were, briefly, these. Compulsive 
drives are specifically neurotic; they are born of feelings of isolation, helpless- 
ness, fear, and hostility, and represent ways of coping with the world despite 
these feelings; they aim primarily not at satisfaction but at safety; their compul- 
sive character is due to the anxiety lurking behind them. Two of these drives- 
neurotic cravings for affection and for power— stood out at first in clear relief 
and were presented in detail in The Neurotic Personality. (4, p. 11) 

Later books presented increasingly sophisticated attempts to de- 
lineate the neurotic character structure. Homey has listed many types, 
trends, and conflicting attitudes to this end. All of these constructs 
concern the individual's reactions to others. At the time of her death, 
Horney's systematizing efforts were far from completed. The shifts 
in her flexible development have created the appearance of a brilliant 
disorganization. An over-all survey of her publications, however, 
reveals an internal consistency and a steady progress towards increas- 
ingly complex organizing principles. 

Erich Fromm, like Horney, places the causative factor of neurosis 
in the family, which is seen as the basic "agency" of enculturation. 
Suppressive or hostile parents create the destructive feelings of power- 
lessness and isolation. Human relations and not instinctual pressure 
thus create personality. "Man's nature, his passions, and anxieties are a 


cultural product; as a matter of fact man himself is the most important 
creation and achievement of the continuous human effort, the record 
of which we call history." (3, p. 11) 

Fromm's theories of character are based on the ways in which the 
individual "relates" to his world. He has listed four neurotic mech- 
anisms for "escaping" insecurity (masochism, sadism, destructiveness, 
and automaton conformity) and five character types (receptive, 
hoarding, marketing, exploitive, and productive). All of these are 
directly interpersonal. Fromm's major concern and greatest contribu- 
tion lies not in the area of systematization, but rather in the philosophic 
backgrounds he has provided for the study of personality. The nine- 
teenth century mechanistic pessimism of Freud, clearly inadequate for 
a science of human nature, has received a thoughtful, gentle, and imag- 
inative revision by Erich Fromm. 

Harry Stack Sullivan's most dramatic accomplishment was the 
assertion, which I believe he has demonstrated, that "psychiatry is the 
study of processes that involve or go on between people. The field of 
psychiatry is the field of interpersonal relations under any and all 
circumstances in which these relationships exist." (5, pp. 4-5) Sul- 
livan's most valuable achievement is his demonstration of the "fabu- 
lously more complicated" nature of interpersonal actions and percep- 
tions, and the introduction of observational methods and attitudes for 
making "objective contact with another individual." 

The research and the theories presented in this book are based on 
the writings of Sullivan, and are in some sense an attempt to extend 
them. Although Sullivan's subtle and complex ideas do not summarize 
readily, a brief survey is in order. 

The motive force of personality, for Sullivan as for Horney and 
Fromm, is the avoidance of anxiety. Anxiety, for all three, is an inter- 
personal phenomenon. For Horney it involves the feelings of help- 
lessness and danger; for Fromm, isolation and weakness; for Sullivan, 
loss of self-esteem. Anxiety is interpersonal because it is rooted in the 
dreaded expectation of derogation and rejection by others (or by one- 
self) . The human being is rarely or never free from some interpersonal 
tension; what he does or thinks is generally related to the estimation 
of others. For this reason the motivating principle of behavior is more 
accurately seen as "anxiety reduction" — the avoidance of the greater 
nnxiety and the selection of the lesser anxiety. This is an important 
point to note, because, as we shall see when we deal with interpersonal 
reflexes, it helps explain some of the paradoxical self-punitive behaviors 
by means of which individuals appear to make themselves unhappy. 

Personality is, according to Sullivan, the "relatively enduring pat- 
tern of recurring interpersonal situations which characterize a human 


life." To understand a person is to have knowledge of the inter- 
personal techniques that he employs to avoid or minimize anxiety and 
of the consistent pattern of relationships that he integrates as a result 
of these techniques. 

It is important to note that interpersonal behavior refers to private 
perceptions, conscious reports, symbolic and unwitting expressions, 
as well as to overt actions. 

Another crucial difference between Sullivan's conceptions and the 
Freudian is worth comment. According to the orthodox Freudian, 
that which is warded off from consciousness is the instinctual impulse 
or its disturbing derivatives. According to Sullivan, those things which 
are selectively kept from awareness are interpersonal processes, or 
potentialities, or interpersonal feelings which are anxiety-arousing. 

The self-dynamism is created by anxiety, being the system of 
anxiety-diminishing behavior characteristic of the developing indi- 
vidual. SulUvan has distinguished three modes of experience which 
have important implications: the prototaxic, undifferentiated, un- 
verbalized experiences of early infancy; the parataxic, which includes 
private, unwitting personifications of the self or eidetic others; and 
the syntaxic. The latter mode is defined by the "extent that observa- 
tion, analysis, and the eduction of relations is subjected to consensual 
validation 'with others.' . . ." Consensual validation, a concept with 
rich empirical meaning, is the "degree of approximate agreement with 
a significant other person or persons which permits fairly exact com- 
munication by speech or otherwise, and the drawing of generally 
useful inferences about the action and thought of the other." (6, 
p. 177) When two people in an interaction situation are consensually 
agreed on the basic premises upon which the relationship rests, and 
when they concur in their pertinent perceptions of self and each other, 
then they are communicating in the syntaxic mode. This kind of 
honesty between persons is not a common phenomenon. Its experi- 
ence can be unbearably painful due to the anxiety it evokes. 

The discussion so far has carried us with hazardous speed and 
brevity through those conceptions of Harry Stack Sullivan which are 
most appropriate to the purposes of this volume. We leave without 
any description a host of strikingly original theories — on interview 
tactics, on obsessional and schizophrenic states, on the six epochs of 
personality development, on dissociative and selective inattention, to 
name a few. 

The weakest links in Sullivan's strong conceptual chain are the 
systematic. His publications up to the present (including posthumous 
volumes) have broken new theoretical ground that has not been sown 
or harvested. He presents an approach but not a methodology. He 


convincingly buries the much-berated remains of descriptive, Kraepe- 
linian, and negatively-value-toned psychiatry, but provides no sub- 
stitute classification system. The carefully worked-out categories he 
presented — experience modes, developmental epochs, self-dynamisms 
— are far from the minimum required for a science of personality. 

Sullivan provides an attitude (humility) and an approach (par- 
ticipant observation), but not a methodology for the science to which 
he was dedicated. His formal notational structure is disappointingly 
disorganized and incomplete. 

The Theories of Erik H. Erikson 

In the preceding section we have considered the contributions of 
three personality theorists who have abandoned the libido conception 
and espoused a social or interpersonal point of view. Horney, Fromm, 
and Sullivan do not deny the importance of sexual and biological 
factors. Sullivan, for example, divides human performance into two 
categories based on the "end states" or goals which are involved. The 
first involves "satisfactions," by which Sullivan denotes bodily activ- 
ities. The second end state is "security," which refers to the inter- 
personal or cultural responses. Having paid his respects to the biologi- 
cal facet of human behavior, Sullivan went on to focus almost 
exclusively on security operations and the social dimension of behavior. 

In contrast to the antilibido theorists mentioned above, there is a 
fourth social system of personality which attempts to develop ego, 
cultural, and interpersonal conceptions within the basic framework of 
the Freudian psychosexual theory. This is the work of Erik H. 
Erikson. (1) 

Erikson includes in his systematic writings three personality proc- 
esses, the somatic, the ego, and the societal. He demonstrates (by 
means of a brilliant marshaling of clinical material) that a human 
event cannot be understood unless the relativity of these three factors 
is grasped. 

We study individual human crises by becoming therapeutically involved in 
them. In doing so, we find that the three processes mentioned are three aspects 
of one process— i.e., human life, both words being equally emphasized. Somatic 
tension, individual anxiety, and group panic, then, are only different ways in 
which human anxiety presents itself to different methods of investigation. . . . 
As we review each relevant item in a given case, we cannot escape the convic- 
tion that the meaning of an item which may be "located" in one of the three 
processes is co-determined by its meaning in the other two. An item in one 
process gains relevance by giving significance to and receiving significance from 
Items in the others. Gradually, I hope, we may find better words for this 
relativity in human existence— &s we shall tentatively call what we wish to 
demonstrate. (1, p. 33) 


Erikson has made the most sophisticated and successful attempt to 
integrate historical, sociological, anthropological, and biological data 
into a personality system. He takes for his model of individual char- 
acter structure the Freudian psychosexual theory to which he has 
added an interpersonal terminology. His commitment to the biology 
of the libido theory is stated quite directly. "It will seem to some that 
I am abandoning this point of view [i.e., the importance of interper- 
sonal regulation patterns] as I now proceed to review the whole field 
of what Freud called pregenital stages and erotogenic zones in child- 
hood and attempt to build a bridge from clinical experience to observa- 
tions on societies. For I will again speak of biologically given poten- 
tialities which develop with the child's organism. I do not think that 
psychoanalysis can remain a workable system of inquiry without its 
basic biological formulations, much as they may need reconsidera- 
tion." (l,p. 65) 

Erikson has expanded and "socialized" the Freudian timetable of 
psychosexual adjustment by means of two ingenious systematic 
devices — his conceptions of zones, modes, and modalities and his 
theory of the eight stages of man's psychological development. 

Erikson focuses on three major zones of psychosexual activity — 
oral, anal, and genital. He then defines five modes of approach or basic 
interpersonal vectors which can be expressed by any organ zone. 
These are incorporative 1 (sucking), incorporative 2 (biting), reten- 
tive, eliminative, and mtrusive. A matrix of the combination of zones 
and modes provides a neat device for classifying the fixations, regres- 
sions, and sequences of normal development. 

An even more original conversion of Freudian developmental 
theory to interpersonal language is accomplished by Erikson by means 
of his eight stages of human emotional growth. This is a "list of ego 
qualities — criteria by which the individual demonstrates that his ego, 
at a given stage, is strong enough to integrate the timetable of the 
organism with the structure of social institutions." Erikson holds that 
the individual at each sequential stage of life meets a nuclear conflict, 
the solution for which "is based on the integration of the earlier ones." 

The eight nuclear conflicts according to Erikson are: 

Stage of Life Cycle Nuclear Conflict 

Oral Sensory Trust vs. Mistrust 

Muscular— Anal Autonomy vs. Shame, Doubt 

Locomotor— Genital Initiative vs. Guilt 

Latency Industry vs. Inferiority 

Puberty and Adolescence Identity vs. Role Diffusion 

Young Adulthood Intimacy vs. Isolation 

Adulthood Generativity vs. Stagnation 

Maturity Integrity vs. Disgust, Despair 


Erikson's commitment to an interpersonal and cultural point of 
view stands out clearly in this list of ego qualities. The extraordinary- 
power and significance of Erikson's work is this: he has developed a 
social conception of human nature which certainly equals in com- 
plexity those of Fromm and Horney — and he has done it within the 
broad framework of the Freudian libido theory. He seems to have 
succeeded in his attempt to build a bridge between psychosexual 
theory and social behavior, and has additionally erected a system 
which is eminently heuristic. 

There is therefore considerable justification for considering Erikson 
as the first major psychoanalytic systematist since Freud. He has, it 
must be noted, surpassed Sullivan on his own home ground by pre- 
senting a developmental timetable which lists sixteen interpersonal 
resolutions. This provides us with an impressive list of interpersonal 
variables lacking in the writings of the less systematic Sullivan. 

The interpersonal system of personality to be presented in this 
book has leaned heavily upon the conceptions of Erik H. Erikson. Our 
classification of interpersonal behavior bears the unmistakable mark of 
Erikson's theory. We have been able to utilize only a fragment of his 
system. This is because Erikson's writings range deep and wide — deep 
into childhood and wide into society. Our own purpose and efforts 
are much more restricted since we have attempted simply to develop 
an objective, functional system for predicting the behavior of adult 
patients in the psychiatric clinic. 

Interpersonal Behavior Defines the Most Important 
Dimension of Personality 

In the preceding sections we have presented a definition of inter- 
personal behavior and have compared several approaches to human 
nature in the light of their social orientations. The assertion was made 
that the interpersonal can from this point of view be considered the 
most crucial and functionally important dimension of personality. 

First, from the broader theoretical frame of reference, interpersonal 
behavior is crucial to the survival of the human being. From a second 
(and much more parochial) point of view, interpersonal behavior is 
the aspect of personality that is most functionally relevant to the 
clinician. Some justification for the first of these assertions will be 
discussed in the next section. The usefulness of an interpersonal 
theory in clinical practice will be considered in Chapter 5. 

Interpersonal Behavior and Biological Survival 

From the standpoint of human survival, social role and social ad- 
justment comprise the most important dimension of personality. This 


is because of the unique biological and cultural aspects of human devel- 
opment and maturity. 

One of the major differences between man and the other animal 
species is his long and helpless infancy. Depending on the complexity 
of the culture, it takes from 12 to 25 years for a human being to attain 
developmental maturity. This long period of childhood and adoles- 
cence involves a dependence on other human beings for nourishment, 
shelter, and security. Many animal species, on the contrary, are ready 
to undertake complete responsibility for their own survival at birth, 
or shortly thereafter. In these cases instinctual methods of locomotion, 
food collection, and self-protection take over immediately. Rigidly 
built-in patterns of response are vital to their early self-sufficiency. 
Automatic physiological responses are the key to life for these infra- 
human organisms. 

The case of man is quite different. The human infant has limited 
physical capacity and few automatic behavior sequences for dealing 
directly with the physical environment. From the moment of birth, 
survival depends on the adequacy of interpersonal relationships. The 
water, warmth, and milk upon which the infant's life depends come 
from others. These primitive, basic transactions which the neonate 
carries on with others are, we are told, not rigidly fixed patterns. A 
variety of early parental response exists, and this is matched by a varia- 
tion in neonate behavior. Several experts in this field (Sullivan, Klein, 
Erikson, Ribble, Spitz) have claimed that the roots of personality are 
to be found in the earliest mother-child interactions. This claim is not 
surprising when we recall that a raw, intense, basic anxiety (concerned 
with the maintenance of life itself) may be felt by the neonate. And 
this anxiety is dealt with (partially or completely, carelessly or lov- 
ingly, calmly or nervously) by the mothering-one. The earliest 
kind of survival anxiety is, therefore, handled by interpersonal, social 

From the standpoint of physiology the human infant is not much 
different from any young mammal. From the standpoint of per- 
sonahty psychology, however, the human being at birth is an extraor- 
dinarily plastic, germinal nucleus with infinite potentialities for 
eventual differentiation. It might be said that any neonate is a potential 
president, priest, poet, or psychotic. PersonaUty psychology is con- 
cerned with the events and behaviors which determine the emotional 
and social development of the individual. The most important factors 
which account for the wide varieties of behavior characteristic of the 
human being are the interpersonal security operations which he 
develops and the social relationships (real and fantasied) which he 
integrates with others. 


We have pointed to the crucial influence of the earliest social trans- 
actions between mother and child — crucial because of the survival 
anxiety involved and because of the complete dependence of the 

As the child grows, the primacy of interpersonal relationships does 
not lessen greatly. A seven-year-old child has developed many motoric 
patterns for self-protection, but on the hypothetical desert island or in 
any societal context we cannot credit him with survival self- 

The human being maintains existence by virtue of the long period 
of parental protection during which he assimilates the complicated 
cultural wisdom necessary for survival. This process of slow, and 
often painful, learning is intensely interpersonal. 

Even at maturity survival rests upon successful interpersonal pat- 
terns. The mutual dependence of mankind is inevitable. Whether we 
exist in a primitive tribe, a dictatorship, or an industrial democracy, 
the key to human life lies in the adequacy of social interaction. Even 
the rare test case of a hermit falls within the limits of this generaliza- 
tion, since this adjustment technique always involves intense and often 
bitter "withdrawal" from others, and is one pattern of interpersonal 
reactivity. The extent to which we autoniatically and implicitly 
demonstrate patterns of cooperation and submission to social demands 
— even in the most democratic society — is quite striking. Failure to do 
so invites such real or fantasied threats to life that we automatically 
commit ourselves in countless ways to the interpersonal pressure of 
parents, societies, and contemporaries. 

Anxiety Motivates Interpersonal Behavior 

The preceding section is intended to justify the statement that inter- 
personal behavior has a basic survival function. The fear of inter- 
personal disaster is rooted in a fear of destruction or abandonment. 
The organism has hundreds of physiological functions by which de- 
struction is warded off and life preserved. The individual develops, 
in addition, numerous emotional responses which, in their origins, are 
concerned with survival. . 

The psychological expression of the survival drive of evolution 
theory is anxiety. Primal anxiety is the fear of abandonment.^ As the 
child begins to develop, this becomes a fear of rejection and social 
disapproval. Mankind's social interdependence means that extreme 

* In the first version of this manuscript this sentence read, "Primal anxiety is the 
fear of death." The revision vi^as made at the suggestion of Harvey Powelson, M.D., 
who pointed out that death is a sophisticated, complex concept which an infant or 
young child has not mastered. 



derogation on the part of crucial others can lead to destruction. The 
behaviors by which the child avoids derogation are called security 
operations. They assure him of the approval and social security which 
reduce his anxiety. 

As the individual develops, further complications ensue. Self- 
esteem becomes a factor which is equal to, or greater than, the overt 
esteem of others.^ 

The role of anxiety in the development of human personality is 
central, and it is intricate beyond our understanding. Although rooted 
genetically in the fear of death, anxiety (i.e., the fear of disapproval) 
is clearly stronger in the case of the adult than the fear of death. There 
are countless examples of human beings choosing to face and accept 
destruction rather than face anxiety and the loss of self-esteem. 
Suicide is one of many such examples. 

Another complication which must be considered in understanding 
the effects of anxiety involves the multilevel organization of behaviors 
for warding off anxiety. A large percentage of any population, for 
example, develops security operations which entail overt self-efface- 
ment, self-derogation, and the provocation of actual contempt and 
disapproval from others. These overt self-derogations, which seem to 
contradict our theory of anxiety, can be understood by means of a 
multilevel analysis. They are inevitably related to private feelings of 
uniqueness or secret consolations. They serve to protect inner feelings 
of pride and self-enhancement. 

This book and the system of personality which it describes is con- 
cerned with a multilevel investigation of human security operations. 
We have taken as our task the definition, classification, and measure- 
ment of interpersonal behavior (at several levels). We view the inter- 
personal behavior of an individual as the machinery by means of which 
he wards off anxiety and maintains a multilevel balance of self- 

The conceptual model of personality which we are developing 
exposes one area of human behavior to study. This is the interpersonal 
dimension. The theoretical system is based on one assumption about 
the motivation of emotional behavior. This has been formalized as 

First working principle: Personality is the multilevel pattern of 
interpersonal responses (overt, conscious, or private) expressed by the 
individual. Interpersonal behavior is aimed at reducing anxiety. All 

^The complexity of the processes of identification and introj action make this com- 
parison redundant and probably meaningless. There is good reason to believe that self- 
esteem is usually or always based on values which are taken from others. Thus self- 
esteem can be considered an indirect form of approval of crucial others. 


the social, emotional, interpersonal activities of an individual can be 
understood as attempts to avoid anxiety or to establish and maintain 


1. Erikson, E. H. Childhood and society. New York: Norton, 1950. 

2. Fenichel, O. Psychoanalytic theory of neurosis. New York: Norton, 1945. 

3. Fromm, E. Escape from freedom. New York: Rinehan, 1947. 

4. HoRNEY, Karen. Our inner conflicts. New York: Norton, 1945. 

5. Sullivan, H. S. Conceptions of modem psychiatry. Washington, D.C.: The Wil- 
liam Alanson White Psychiatric Foundation, 1947. 

6. SuLUVAN, H. S. Multidisciplined coordination of interpersonal data. In S. S 
Sargent and Marian W. Smith (eds.). Culture and personality. New York: 
Viking Fund, 1949. 

Adjustment-Maladjustment Factors in 
Personality Theory 

It is the theme of this chapter that personality theories should hold for 
adjustive and maladjustive behaviors, that normality and abnormality 
should be defined as different points on the same measurement con- 
tinuum, and that the conceptual terminology of personality should 
therefore include the entire adjustive range of human activity. Few 
theories do this. Most are oriented toward abnormal or neurotic 
behaviors. Most diagnostic systems have few terms for conceptualiz- 
ing adaptive behavior, which is described in vague generalities or in 
terms of the absence of pathology. 

This is an unfortunate state of affairs. It reflects an undeliberate 
but significant depreciation of human nature. In addition, this pathol- 
ogy error tends to distort our theories of personality by placing a 
disproportionate emphasis on certain limited types of maladjustment, 

A science of malfunction cannot precede a science of function. 
Therapeutic tactics can break new ground, but scientific and theoretic 
progress depends upon the development of the principles of normal 
adjustment. The fact that psychiatric theories of personality have been 
based on clinical experiences has led to some curiously one-sided 
conceptualizations. Psychiatry, however, cannot be wholly blamed 
for these restrictions, which, as we shall see, spring from a marked 
asymmetry in the ethical evaluations of varying interpersonal themes 
in our Western culture. 

Before approaching the definitions of adjustment-maladjustment 
we shall review psychiatry's overemphasis on the abnormal, and we 
shall consider some causes and implications of this pathology error. 

Psychiatric Theories Are Oriented Towards Pathology 

A history of man's conception of his own nature has yet to be 
written. When our systematic knowledge of human expressive be- 



havior is more advanced, it will be possible to study the literary and 
historical documents of the past, and to determine the expressed and 
implied views of personality that determined the behavior of our 
ancestors. One tentative generalization — basic to the theory of this 
book — may be helpful in surveying the changing conceptions of 
human nature. This concerns the Locus of Responsibility for human 
behavior. There seems to be a consistent tendency in the development 
of psychological knowledge to move the causative factor of human 
behavior from external to internal forces. This is clearly reflected in 
the changes in the theoretical explanations of abnormal or maladjus- 
tive behavior. 

We are told that success or failure appeared, to the ancients, to be 
controlled by the immutable and mysterious powers of nature. Sun, 
seed, and storm were fearful forces — completely inexplicable. Man's 
survival responses appeared by comparison quite meaningless. The 
shift of causative principles to anthropomorphic gods made human 
behavior somewhat more important. The notion that man can move 
the gods by propitiation, obedience, or defiance considerably human- 
izes the causative sequence. 

This conception which lasted from the Greek civilization through 
to the nineteenth century (and which still is maintained by a large 
majority of individuals living today) defines personality aberration as a 
religious phenomenon. Maladjustment is a mark of omnipotent inter- 
vention, generally indicating a sinful nature. The maladjusted person 
is isolated, overtly punished, or covertly rejected. The error is man's 
and the power is the god's. 

The theories of descriptive psychiatry which emphasized constitu- 
tional morbidity, although they had the ring of scientific objectivity, 
were still very crude conceptions. They were abysmally inferior 
to the insights of the artistic geniuses who preceded them by several 
centuries. Shakespeare, for example, progressed much further from 
the Greek mythology than the average hospital psychiatrist of the 
early 1900's. On the other hand, in the strictest sense of dramatic 
motivation, Oedipus was a morbidly predisposed type — since no 
choice is given him at any point to reverse his awful destiny. This 
type of psychological explanation is quite congenial to pre-Freudian 
psychiatry. Change a few mythological terms and Oedipus is an 
acceptable case history from the textbook of the nineteenth century 
alienist. When we compare this predestined helplessness with the 
self-imposed conflicts of Shakespeare's characters the descriptive 
psychiatrist comes off badly. Although Elizabethan theories of human 
destiny involved chance and fortuitous influences (the wheel of fate), 
still the reader is impressed by the implication that the poet's heroes 


court their tragic ends because of their own greed, ambition, indeci- 
sion, and shallowness. The causative agency has moved from the ex- 
ternal and immovable force to the partial responsibility of the hero for 
his own self-created destiny. 

The notion that human nature and the individual's fate are deter- 
mined by his own (conscious or unconscious) decisions and solutions 
is brilliantly illustrated by Marcel Proust. In Remembrance of Things 
Past he describes how his hero deliberately trains and provokes his 
parents to accept him as a neurotic child. In the following episode he 
literally creates his own maladjustment and develops the weak and 
asocial role he is to maintain in his future life. His parents agree, 
" 'It is his nerves . . .' . And thus for the first time my unhappiness 
was regarded no longer as a fault for which I must be punished, but as 
an involuntary evil which has been officially recognized, a nervous 
condition for which I was in no way responsible: I had the consola- 
tion that I need no longer mingle apprehensive scruples with the bit- 
terness of my tears; I could weep henceforth without sin." 

The narrator recognizes, however, that his neurosis is not "involun- 
tary," but rather a purposive, victorious interpersonal maneuver. He 
has unconsciously selected nervousness as a security operation. The 
narrator then goes on to say, "I ought then to be happy; I was not. It 
struck me that my mother had just made a first concession which must 
have been painful to her, that it was a first step down from the ideal 
she had formed for me, and that for the first time she, with all her 
courage, had to confess herself beaten. It struck me that if I had just 
scored a victory it was over her; that I had succeeded, as sickness or 
sorrow or age might have succeeded, in relaxing her will, in altering 
her judgment; that this evening opened a new era, must remain a black 
date in the calendar." (4, p. 49) 

Freudian Theory of Normality 

With the Freudian theory, psychology begins to catch up with 
the intuitions of literature. Man's character, his responses and solu- 
tions to the overwhelming conflicts of life are brought into focus. 
While man has a choice of reactions which bring relative amounts of 
temporary security, the balance, according to Freud, is still on the side 
of the native, instinctual endowment. The doctrine of instincts em- 
phasizes the inevitable pressure of drives external to the ego. In early 
psychoanalytic theory it is libidinal drive that is basic, inborn, con- 
stant, and, in the final sense, victorious. The adaptive forces are 
acquired, inconstant, variable, and, in the final sense, secondary. In 
fact, the ego functions, defense mechanisms, and character traits were 
sometimes interpreted as neurotic solutions. 


By building his logical notational structure on the "id" instincts, 
Freud was making a formal decision, and not an empirical discovery. 
Impressed by the new insights he obtained into the antisocial impulses, 
it was natural for Freud to base his theory on that aspect of human 
motivation. It is generally accepted that later developments in psycho- 
analysis have reversed this trend, and have placed more emphasis on 
the ego, studying its structure, function, and the multiplex variety of 
its processes. But it is also commonly known that early psychoanalytic 
terminology tends to lack terms for describing adjustive behavior and 
normal processes. The conceptual contributions of Erikson have 
competently filled in this gap in the psychoanalytic nosology. 

The psychoanalytic theory of personality, which is by far the 
most complete and complex theory, is based on the statistically narrow, 
neurotic extreme of the general population in two or three Occidental 
countries. As we shall see subsequently, there is good evidence to 
suggest that early psychiatric and psychoanalytic theory was based 
on less than one half of the range of this maladjustive extreme, and 
that perhaps 50 per cent of neurotic solutions remained largely un- 

The curious phenomenon of a massive theoretical structure erected 
on an emaciated sample of subjects is, I believe, due to two basic 
factors, one logical and one empirical. Freud's formal choice in em- 
phasizing the destructive strivings is historically comprehensible, and 
no detraction from his creative genius. 

The empirical factor, as I have suggested, refers to the narrow 
range of individuals whose neurosis is such as to lead them to submit 
to the singular and rather implausible process of psychoanalysis (cf. 
Chapter 12). 

Jung's Emphasis on Adaptive Behavior 

The Jungian school of analytic psychology produced several im- 
portant revisions of Freudian concepts. Most of its unique contribu- 
tions are refinements and extensions of Freudian theory. To the extent 
that any cognitive issue was involved, we can say that the Zurich 
group split off from Vienna when Jung rejected the narrow sexual 
interpretation of libidinal energy. By broadening the meaning of this 
basic impulse, Jung and his followers have made it general and vague, 
and thus relegated it to a secondary theoretical position. This indirect 
shelving of the libido theory can be taken as an unpremeditated, but 
vital, aspect of the Jungian position. Other revisions pertinent to this 
discussion include theories of functions, neurosis, and unconscious 


The Jungian functions — extroversion, intuition, thinking, etc. — are 
seen as important, but not necessarily negative, psychological mech- 
anisms. They are pathological only when rigidly misused or when 
completely repressed. For the most part, when a Jungian diagnostician 
calls the subject "introverted" he is not making a value judgment; he 
implies only that this is an important way in which the patient handles 
experience and its conflicts, and it may or may not be necessary to 
modify its use. 

It follows, then, that the Jungians do not see character distortions as 
pathological fixations or regressions to inevitable infantile stages. They 
describe neurosis as a partial solution to life's dilemmas — a construc- 
tive mobilizing of "psychic" resources against real or imagined threats. 
They might say of the neurotic pattern, "This is a good try, perhaps 
the best you could do under those circumstances. Now let's see what 
the results of these solutions have been and what other possibilities for 
resolution we can discover." 

This approach has much to recommend it. It is very congenial to 
the current medical conception which defines disease, not as an un- 
fortunate falling ill, but as a complicated interaction between one net- 
work of adaptive responses and another network of threatening events. 

Another, and perhaps the greatest, advantage of the Jungian system 
is the conception of unconscious motivations as valuable, undiscovered 
potentials of the self, rather than as destructive impulses. Bateson has 
appraised the Jungian viewpoint as more consistent with the prin- 
ciples of communication theory. He points out that 

the Freudian ambition to substitute ego for id or to include the id within the 
scope of the ego, sounds to Jungians like advocating manipulative and conscious 
control of the foreign body. In reply to this they would urge merely the 
acceptance— even the joyful acceptance— of the fact that the foreign body 
though always and inevitably unconscious is really a part of the self and the self 
a part of it— the collective unconscious being imagined to be in some sense 
greater than the self. (5, p. 264) 

With this background it becomes clear that the Jungian theory, 
although based on and indebted to the work of Freud, has made certain 
advances toward a balanced conception of normahty-abnormality 
factors. Shifting the stress from infantile strivings to the selecting and 
adapting functions of response helps to free psychology from fatalistic 
themes which have limited man's view of human nature from 
Sophocles' time through Freud's. 

Jungian theories have contributed, often indirectly, to four promis- 
ing notions. First they bring us closer to the development of a 
normality-abnormality continuum, which makes neurosis not a quali- 


tatively different phenomenon. They help us see the interaction be- 
tween biological-cultural pressures and the adaptive-maladaptive 
responses of the individual. They emphasize the "circular or reticu- 
late" equilibrium of different levels of personality rather than the 
one-sided organization for warding off unconscious motivations. 
Finally, they are, perhaps, the first to introduce the far-reaching idea 
that unconscious or repressed motives can be positive, constructive 
potentials, and are not necessarily negative. 

Many of these doctrines were only implicit in Jung's writings, and 
credit for their informal, undramatic development must be assigned 
to certain American analytic psychiatrists, in particular, Joseph Wheel- 
wright and Joseph Henderson. 

Homey and Fromm on Normality 

This general tendency to focus upon adjustive behavior has been 
given articulate expression by psychiatric systemists who have em- 
phasized the cultural dimension of personality. When Horney and 
Fromm substitute cultural factors for instinctual pressure in the 
causative formula, they bring about drastic revision in attitudes 
toward mental health and disease. In the first place, the sexual and 
aggressive instincts — defined by Freud as universal, immutable, and 
antisocial — tend to taint all men with a new form of original sin. The 
culture concept is much more flexible. It gives man, or sohie men, a 
halfway chance because of the wide variation in social environments 
and cultural pressures. 

Thus the diagnostician's causative questions become: "What were 
the set of biological, familial, social, and cultural pressures which this 
patient faced, and what was the particular network of responses by 
which he dealt with them?" The issue of normality-abnormality takes 
on new meaning in this context. A survey of the publications of 
Fromm or Horney will reveal the extent to which these authors are 
concerned with the individual's attempts to solve his conflicts. We 
have in the previous chapter cited a partial list of some dozen mech- 
anisms, escapes, and trends described by these two theorists. Over 
and over again they emphasize the response of the patient to the en- 
vironment, and his interactions with it. Their interest in pathology is 
always hnked to the underlying notion that neurosis is acquired by 
and through the individual's reactions to social stress, and the sub- 
sidiary idea that it can be "cured" by shifting one's reactions to stress 
in the future. 

Basic and implicit to the theories of both is the theme that mal- 
adjustment is different in degree, and not in kind, from the so-called 
norm. Fromm states this clearly, 


The phenomena which we observe in the neurotic person are in principle not 
different from those we find in the normal. They are only more accentuated, 
clear-cut, and frequently more accessible to the awareness of the neurotic person 
than they are in the normal who is not aware of any personal problem which 
warrants study. (1, p. 17) 

Sullivan and the Concept of Normality 

Within the framework of a brief historical review we have been 
selecting several themes which comprise the message of this chapter. 
These include the qualitative similarity of normality-abnormality, the 
locus of responsibility assigned to the individual's behavior, rather than 
to fatalistic forces, and the necessity to take into account the multi- 
level nature of human potentialities. These concepts, which are im- 
plicit in the development of psychoanalytic theory during the last 
fifty years, appear over and over again in the writings of Sullivan. 
This theorist, we recall, holds that the self is formed through the 
child's sensitivity to approval and disapproval. If we accept this notion 
that personality is determined by interpersonal anxiety we have closed 
the qualitative gap between normal and abnormal. ''Everything that 
can be found in mental disorder can be found in anyone, but the 
accent, the prominence, the misuse, of that which is found in the 
mental patient, is more or less characteristic." (3, p. 77) With this re- 
mark SuUivan advances the concept of the continuity of normal 
and abnormal human behavior which developed from the orig- 
inal Jungian protest. Listing neurotic and normal behavior along a 
relativistic continuum is a humanistic trend, which results in changing 
techniques in psychotherapy. Moreover, it lends itself more directly 
to scientific procedures, since probability laws become considerably 
more feasible. The pathology error in psychiatric thinking led to 
theorems that were based on neurotic behavior, and which had little 
to say about normal functioning. By concentrating on the processes of 
adaptation in their successful and unsuccessful forms the stage is set for 
many new personality systems which will hold for all human behavior. 

Emphasis on Adaptive Responses Leads 

to a Neutral Conceptio?i of Human Nature 

To insist that psychology focus on man's executive, adaptive reac- 
tions — in their adjustive flexibility as well as their maladjustive 
extreme — is not to argue for a bright-eyed optimistic view of the 
human situation. In many ways it is much kinder to inform a fellow 
human being that his misery or failure is due to divine direction, 
inherited disposition, or biological destiny. We remember that 
Sophocles, while plunging Oedipus into the depth of despair, never 
forced him to express man's most poignant lament, "I could have done 


differently." His fate was always in the hands of the gods. The 
responsibility for human destiny is thereby transferred to external 
forces. This view relieves man of the obligation to effect change, 
which is assigned to omnipotent powers among whom later generations 
have included the physician. This is probably the easiest and most 
comfortable conception of human nature. 

When we interpret adjustment in terms of the individual's own 
responses, rigid solutions, and escape mechanisms, we present our 
fellow-sufferers with an ambiguous gift. Two rather staggering im- 
plications accompany this conception. Neither are particularly optim- 
istic. The first is, "You must accept the blame or credit for your 
present situation; you, and not your rejecting parents, your race, 
your instinctual heritage, your drunken husband, but your own pat- 
tern of repetitive and self-limiting responses created it." To this grim 
frankness we must add the corollary, "To you, therefore, is given the 
power to change your situation. . . . it is impossible and unnecessary 
to change your childhood, the society in which you live, your skin 
color, your biological make-up, or your spouse — what is required is a 
change in your inaccurate perceptions and rigid reactions." 

When we replace immutable external forces with self-determinism, 
we invite the individual to accept a most lonely and frightening 
power which, as Fromm has pointed out, none of us are well trained 
to assume. This is, of course, neither an optimistic nor a pessimistic 
point of view, being rather the neutral realistic statement of the 
reciprocal principles of social interaction and self-determination. 

Symptom and Character 

The changing approaches to personality just described have re- 
sulted in an additional clarifying abstraction which is very pertinent to 
the conception of neurosis. This is the distinction between symptom 
and character. 

As used in this context, the term character refers to the personality 
— the durable, multiple-level pattern of interpersonal tendencies or- 
ganized into stable or unstable equilibria. This complex organization 
of perception and action is a logical notational structure by which we 
conceptualize the anxiety-reducing operations of the individual. It is 
the theoretical and linguistic device by which we summarize our 
knowledge of a human being. The character structure, as the sum 
total of an individual's interpersonal behavior, is the psychologist's 
shorthand for the social human being. 

A symptom, as succinctly defined by Masserman (2, p. 298), is 
any "overt manifestation of a disease or behavior disorder." It is one 
aspect of the unified network of variables that make up personality. 


and an important aspect in that it indicates an imbalance or malfunc- 
tion in the character structure. A symptom not only tells us that 
something is distorted in the personality, but in the nature of its 
specificity often suggests what kind of a distortion exists. Regardless 
of how centrally painful psychiatric symptoms may be to the patient 
or to his intimates, their meaning, function, and treatment must be 
viewed as one set of factors related to many others in the personality 

To illustrate the distinction between symptom and character, let 
us pose the question, what do we mean by neurosis? Psychiatric text- 
books define neurosis in terms of repetitive, anxiety-driven behavior 
based on internal conflict, and manifesting certain symptomatic ex- 
pressions. This is a broad, inclusive, dictionary-type definition, and a 
pretty good one. It emphasizes not only the external appearance of 
neurosis — the symptoms — but also the underlying character distor- 
tions. Unhappily when the nonanalytic psychiatrist takes off his 
Sunday-best terminology and lists his workday operating diagnostic 
concepts, this nice balance is lost. Most, if not all, of the commonly 
used psychiatric categories — schizoid, depressive, psychopathic, psy- 
chosomatic — are symptom-oriented. They are based on certain ex- 
ternal signs of unsuccessful adaption. In practice, an individual is 
diagnosed as neurotic if he manifests the so-called psychiatric symp- 
toms which are restricted to a certain range of social inefficiencies. 
Most patients come to the psychiatric clinic not expressing dissatis- 
faction with their character, but requesting relief from symptoms. 
The attention of the patient and most preanalytic therapists is 
naturally directed to the painful, and often terrifying external manifes- 
tations of psychiatric distress. This symptom orientation supplies 
another reason why psychiatry and the personality theories it has 
produced have taken on the negativistic, neurosis-bound cast which 
we have called the pathology error. 

The attempt to develop personality theories in the atmosphere of 
the consulting room and clinic has resulted in still another interesting 
limitation. The second half of this compound fallacy is caused by the 
fact that (until the last decade), of all neurotic character types, only 
about one half came in any frequency to seek psychotherapeutic help. 
We can suspect that about 50 per cent of individuals with marked 
character distortions (i.e., one half of the diagnostic continuum) did 
not show up in large numbers in the nineteenth century psychiatric 
office because the very essence of their imbalance tended to push them 
away from dependence, self-revelation, and conforming cooperation. 

The diagnostic chapters of this book will consider this interesting 
phenomenon in some detail. It is pertinent to the argument here to 


point out that a large percentage of the maladjusted population has 
traditionally received little psychiatric attention. They were not 
studied because they did not come for psychiatric help. They did not 
seek therapeutic assistance because the core of their anxiety-reducing 
operations was a compulsive maintenance of povi^er, independence, 
competitiveness, or defiance — interpersonal techniques which pre- 
clude, under ordinary circumstances, the role of a psychiatric patient. 

Working Principle II: Adjustive-Maladjustive 
Personality Variables 

Thus, our personality theories have not only been lopsided in the 
direction of maladjusted rather than normal subjects, but also limited 
by overemphasis on a narrow fragment of the over-all neurotic popu- 
lation. We can now present the second principle upon which the 
interpersonal system is based. 

Second working principle: The variables of a personality system 
should be designed to measure — on the same continuum — the normal, 
adjustive aspects of behavior as ivell as abnormal or pathological 

In validating a system of personality, the procedures of data collec- 
tion should include samples of both adjusted and maladjusted subjects. 
Among the maladjusted there should be proportionate empirical 
attention to those subjects whose anxiety is lessened by rushing- 
into-a-psychiatric-clinic as well as those whose anxiety is dimin- 
ished by a rushing-away-from-the-interpersonal-implications-of-the- 

By basing their conceptions on the human character structure, 
rather than on a fractional segment of symptoms, Erikson, Horney, 
Fromm, and Sullivan have doubled the range of personality types. We 
learn that many apparently successful and socially approved behavior 
extremes — the driving competitor, the overambitious leader, the over- 
popular hero — can be based on imbalanced and neurotic character 
structures. It is easy to add the corollary that many phenomena clas- 
sically considered deeply pathological — mild autistic withdrawals, 
moderate unconventionality, moderate depressed obsessiveness — are 
not severe imbalances but constructive, healthy, and perfectly ac- 
ceptable methods of warding off anxiety. 

Effect of Cultural Values on Theories of Normality 

Fromm speaks in this connection of the difference between per- 
sonal and social maladjustment. Social efficiency manifested by public 
esteem, high income, and feverish productivity may give the appear- 


ance of healthy adjustment at the expense of disequilibrium and in- 
ternal distortion. Social inefficiency, defined in terms of low income, 
nonconformity, modest station, social introversion, and relaxed am- 
bition does not always indicate unhappiness or psychic disturbance. 
Poets have known this for some centuries. 

The basic values of the American middle class, which insidiously 
permeate all of its members, exert their influence on contemporary 
psychiatric theories. It is very easy to identify normality with con- 
ventionality or optimistic, active, responsible independence; and 
neurosis with nonconformity or pessimistic, inactive sensitivity. 

The definition of adjustment is thus complicated by the inevitable 
pressure of value systems: Is it more "normal" to express constructive, 
conjunctive, conventional affiUative feelings? Is it more "abnormal" 
to manifest distrustful, hostile, rebellious behavior? 

The personality theorist need not base his definition on cultural 
values, but it is certainly necessary to take into account the social and 
ethical esteem which attaches to certain popular security operations. 

There are two issues which must be faced — a quantitative and a 
qualitative consideration of adjustment. 

Quantitative Definition of Adjustment 

This book is presenting a system for diagnosing personality which 
strives to be objective and operational. This commits us to a quan- 
titative definition of maladjustment. We set up continua for measur- 
ing or classifying interpersonal behavior in terms of several indices. 
Normality-abnormality is defined in terms of these indices. 

The first of these quantitative scales concerns consistent modera- 
tion versus intensity at any one level of behavior. The former is con- 
sidered adjustive, the second maladjustive. 

The second categorization concerns flexibility versus rigidity at 
any one level of behavior. The former is considered adjustive, the 
latter maladjustive. 

A third quantitative index of normality involves the stability or 
oscillation among different levels of personality. Extreme conflict 
(oscillation) among levels is viewed as maladjustive. So is extreme 
interlevel rigidity, i.e., the same interpersonal operations repeated at 
all levels. Stable or balanced interlevel patterns are seen as adjustive. 

A fourth (and less clear-cut) definition of normality involves 
measurements of accuracy and appropriateness. If behavior is in- 
appropriate, if perceptions are inaccurate, then maladjustment is 

The methodology and specific apphcation of these quantitative in- 
dices will be described in later sections of this book. 


Qualitative Definition of Adjustment 

A second approach to the definition of adjustment and maladjust- 
ment involves a qualitative assessment of behavior. Here we do not 
ask "how much?" or "how rigid?" or "how accurate?" but concen- 
trate on luhat kind of interpersonal behavior. 

The qualitative definition of normality is inextricably rooted in 
value judgments and does not appear to be useful in developing an 
objective diagnostic system. The quantitative concept of adjustment 
is based on the notion of personal adjustment. How balanced, ac- 
curate, adaptable are the security operations? How successful are they 
in warding off anxiety? The qualitative concept is based on social 
adjustment — conformity to cultural stereotypes as to what is normal. 

Let us grant that no human being is perfectly balanced, and that 
everyone has developed modes of dealing with anxiety which em- 
phasize certain interpersonal behaviors and minimize others. The 
qualitative question then becomes: Are there socially preferred kinds 
of security operations? Are there certain modes of response which 
are intrinsically better than others? 

Is conventionality or loving trust, for example, intrinsically more 
adjusted than bitter rebellion? 

There is no answer to these questions. This is a cultural, ethical 
issue. The neutral position of the scientist (which of course is an ideal 
and never an actuality) can be preserved by accepting explicitly 
quantitative definitions of adjustment and avoiding (as far as it is pos- 
sible) the qualitative. 

By way of illustration, let us consider two patients, both of whom 
have intense underlying feelings of despair and a long history of 
deprivation and derogation. One patient reacts to these inner feelings 
and experiences by means of a rigid conventionality and conformity 
to duty. The second patient reacts to the same inner feelings and the 
same unhappy history by means of a rigid rebellion and bitter rejec- 
tion of conventional behavior. 

Assuming the rigidity and intensity of the two security operations 
to be equal, is one more adjustive than the other? A quantitative 
definition would hold that there is no difference. 

A qualitative definition might tend to consider one more normal 
than the other. Certainly, most cultural, ethical values would prefer 
the former conforming, cooperative operations and disapprove of the 
latter. But from the standpoint of the individual and his quest for se- 
curity it will be seen that both may achieve the same amount of self- 
esteem and suffer from the same amount of conflict. They may be 
equally successful in warding off anxiety. 


Large and diverse samples of subjects studied by means of dis- 
ciplined, logical variable systems offer the best protection against one- 
sided success-oriented personality theories. The invaluable assistance 
of formal classification and notational structures in systematizing the 
data of human nature is one of the basic maxims of this book. The 
following chapter is devoted to this topic. The following example will 
serve to illustrate its usefulness in the context of the present discussion 
of social versus personal adjustment. 

Illustration of the Impact of Cultural Values 
on Conceptions of Moral Character 

In the process of developing a systematic list of interpersonal 
variables it is obvious that hostile and affectionate behaviors are among 
the commonly employed means of dealing with others. When we 
apply the principles of the normality-abnormality continuum, it fol- 
lows logically that we must have linguistic terms for describing inter- 
mediate points along the continuum between these two interpersonal 
motives. This is to say, we must measure the moderate-adaptive and 
the intense pathological extremes of each morive. Thus, in devising 
rating scales, diagnostic terms, test check lists and the like, it is formally 
required that we have signs or terms to reflect the adjustively hostile, 
the adjustively affectionate, the maladaptively hostile, and the mal- 
adaptively affectionate. 

When the Kaiser Foundation psychology research project began 
to develop a system of interpersonal variables, a puzzling linguistic 
situation was uncovered. It became clear that the English language — 
whether that of the psychiatrist or that of the general public — has a 
marked imbalance in the number of terms which describe different 
interpersonal themes. There was no trouble in obtaining long columns 
of words describing the positive, socially adaptive expressions of 
friendliness, amiability, love, agreeability, etc. Nor was there difficulty 
in listing maladjustive, pathologically toned denotations of extreme 
hostility, hatred, opposition, rage, etc. It was, however, a tedious task 
to get three or four commonly used words for the concept of adjustive, 
socially approved hostility. Considerable dictionary, thesaurus, and 
literary research uncovered a few such words — frank, blunt, critical — 
but it appears that the English language, and the implicit folk con- 
ceptions of human nature that underly it, pay little attention to the 
theme of appropriate expression of disaffiliative interpersonal behavior. 

Interpersonal check lists were given to large samples of diverse sub- 
jects in order to obtain a balanced variable system and to determine the 
expected frequency of social motivations attributed to self. The logic 
of the personality system and statistical simplicity demanded a balance 


between hostile and friendly terms, but the one-sidedness of the inter- 
personal terminology and conceptualization of Anglo-American cul- 
ture made it necessary to employ such clumsy terms as righteous 
anger, not afraid to be critical, and the like, in order to express the 
theme of adaptive, appropriate hostility. 

When we seek to find terms which express extreme, rigid mal- 
adjustive affectionate behavior, the problem becomes insoluble. 
There are no such simple words in the language. According to our 
linguistic forefathers, the human being cannot be too loving. The 
notion that one can be neurotically or compulsively affiliative is 
literally unthought of. 

In this instance, the logical principles of the normality-abnormality 
continuum of interpersonal behavior and the discipline of a formal 
notational system lead to some interesting semantic, anthropological 
speculations and a further illustration of the one-sided clinical error. 

What Is Adjustment? 

In pointing out the limitations of classical psychiatry, and in advo- 
cating expanded symmetrical, logical principles for dealing with the 
normality-abnormaUty continuum, we have left untouched two vital 
questions: What is normality? What is neurosis? These are crucial 
issues because the theoretical position assumed on these questions is 
inextricably bound to the resulting conceptions of personality organi- 
zation, diagnosis, and therapeutic orientation. 

Horney presents changing and developing definitions of neurosis in 
her different publications. In general, she appears to see normality as 
flexibility, optimal productivity, as well as a relative emancipation from 
anxiety and the conflicts which accompany it. Fromm stresses produc- 
tiveness, responsibility, mature affection, understanding, a rational 
handling of the authority relationship, and "freedom" from irrational 
dependence. Sullivan defines mental health as accurate, mutually re- 
warding interpersonal relationships. All of these authors are aware of 
the effect of the culture on our conception of normality. They point 
out that deviation from the norm must be viewed in the context of the 
social background. When Sullivan ties his most adequate mode of ex- 
perience — the syntaxic — to consensual validation he recognizes cul- 
tural relativity, and holds that a "great deal of most people's syntaxic 
experience is bound by the prescriptions and limitations of the 
culture . . ." 

When we survey these criteria of normality, two thoughts may 
occur. First, they are all partially vahd, in the sense that they refer to 
aspects of adjustive functioning. Second, none of them is complete, 
systematic, or too well organized. Productivity, syntaxic function, 


and achievement of one's potential are broad concepts, admirable 
foundations for a philosophy of human nature, but much too vague 
and general to be used as research and clinical variables. 

From the standpoint of operational measurement, most definitions 
of normality are either too specific, and thus fragmentary, or too 
broad, and thus imprecise. This is because normality cannot be sys- 
tematically defined until a comprehensive system exists for organizing 
the multiplex data of human nature. Personality processes operate at 
many levels and in many forms. The nature of the definition of 
neurosis is always chained to the nature of the system of variables by 
w^hich the theorist classifies human behavior. 

We shall obtain rigorous, logical, complex heuristic definitions of 
adjustment-maladjustment when we are given systematic multilevel 
definitions of human personality. Until then the conception of 
neurosis will reflect the level of personality to which the theorist is 

At this point in the discussion it is appropriate to introduce the 
theory of normality basic to the personality system presented in this 
book. To venture its definition at this early stage of the exposition is a 
hazardous proposition. Since a detailed description of personality 
organization has not been presented, a detailed definition of normality 
is premature. We shall be forced to employ undefined words, refer to 
undefined levels and their undescribed relationships. Fluent expres- 
sion of nonoperationally defined terms is the easiest trap that awaits 
the personality theorist. We shall, with these reservations, present a 
verbal description of normality, at the same time referring the reader 
ahead to the systematic and operationally defined categories which are 
to follow in Chapter 12. 

Adjustment in terms of the over-all personality organization con- 
sists in flexible, balanced, appropriate, accurate interpersonal behavior. 
In terms of the subdivisions of personality — the levels of public inter- 
action, perception, and private symbolism — it consists of appropriate, 
accurate, and balanced interpersonal behavior respectively. When we 
re-examine this definition we shall see that each term has a rigorous 
quantitative meaning — referring to specific, operationally defined 
processes. In the broad scope, we call normality an equilibrium of all 
the levels of personality such that the necessary mild character distor- 
tions at some levels are moderately counterbalanced at other levels. A 
different subdefinition exists for each different level of personality. At 
the level of perception of self or others, accuracy or syntaxic agree- 
ment with consensual perception is a partial index of adjustment. At 
the level of overt interaction, the proportion of flexible interactions 
appropriate to the interpersonal stimulus becomes the index of adjust- 


ment. At the level of indirect, fantasy expression, the breadth of 
symbolic themes and their balance and relationship to the other levels 
provides the ratio of adjustment/ 

The verbal definition of adjustment presented above rests upon one 
basic (philosophic) assumption: survival anxiety as the motivating 
force of interpersonal behavior. This premise shapes the resulting 
theory of normality. It also focuses on certain types of variables 
(interpersonal), and requires certain formal multilevel systems for 
relating these variables. The conception of adjustment-maladjust- 
ment presented in this section, therefore, does not stand as an isolated 
verbal entity. This will become clearer as we examine, in later chap- 
ters, the specific and, in the following chapter, the general principles 
of the system on which it is based. 

' In Chapter 12 operational methods for classifying and diagnosing behavior will be 
presented. This conception of adjustment is based on the notions of moderation, bal- 
ance, and flexibility. In developing objective criteria for measuring these qualities we 
have found ourselves borrowing from certain historical antecedents and rejecting 
others. Moderation and the avoidance of extremes is, of course, the definition of 
adjustment sponsored by Aristotle. Flexibility and the avoidance of narrow, rigid 
forms of adjustment is the Renaissance ideal. The Christian conception of values 
views normality as a victor over man's intrinsic evil nature. This notion is reflected 
in the psychiatric theories of adjustment developed in the nineteentli century. It is a 
curious irony that empirical approaches to the definition of normality find their 
intellectual heritage in the Greek and Renaissance philosophies which are more distant 
in many other respects from the ethos of t%ventieth-century culture. 


1. Fromm, E. Escape from freedom. New York: Rinehart, 1951. 

2. Masserman, J. H. Principles of dynamic psychiatry. Philadelphia: Saunders, 

3. MuLLAHY, p. The theories of Harry Stack Sullivan. In P. Mullahy (ed.). The 
Contributions of Harry Stack Sullivan. New York: Hermitage House, 1952. 

4. Proust, M. Swann's way. Translated by C- K. Scott-MoncriefT. London: Chatto, 
1922. Vol. 1. 

5. RuEscH, J., and G. Bateson. Communication. New York: Norton, 1951. 

Systematizing the Complexity 
of Personality 

That segment of personality which we have selected to systematize 
centers on adjustive and maladjustive interpersonal behavior. Even 
when we narrow our field to the social dimension of personality, the 
systematic task remaining is terribly complicated. The diversity of 
interpersonal behavior covers a wide range. It includes all the things a 
subject does to others at all levels of personality — overtly, symboli- 
cally, and in private perceptions. When we add the parallel behaviors 
of others who do things to the subject we obtain a network of events 
that probably equals in complexity the data of the physical sciences. 
When we consider further the effects of culture, sex difference, and 
the peculiarly self-deceptive nature of emotional data, the enormity of 
the scientific task becomes clear. 

In undertaking this complex mission, personality psychology can, 
fortunately, count on some conceptual assistance — new developments 
in the philosophy of science. In recent years considerable progress 
has been made by a group of logicians and positivist philosophers 
which is directly apphcable to the field of personality. The study of 
human nature can find guide posts in the general principles which 
guide the physical sciences. 

The Basic Conceptual Unit of Personality 

We shall begin by considering a preliminary question. When we 
study the interpersonal behavior of an individual, what is the basic 
datum on which we make our judgments? The first answer to this 
question might be that we employ a variety of behavioral cues: projec- 
tive personality tests, tales of woe from the interview, the angry tones 
of voice, dream texts, and the hke. These are, it is true, the events, but 



they are not the basic data for the study of personality. How can we 
measure these written, oral, and physical expressions in such a way as 
to provide comparative conceptual material? It is possible, but rarely 
feasible, to capture these events by sound and movie equipment. Even 
then we must decide what to do with these unwieldy materials when 
we get them. 

For many years researchers have been working within one or an- 
other of these areas of raw personality data, painfully building up com- 
plex devices for categorizing the different surface types of expression. 
Hundreds of systems for dealing with personality tests have been pub- 
lished. We have learned, to our horror, that it is possible to devise 
measurement scales for each facet of personahty expression. Thus, it 
is possible to have an elaborate continuum for rating each type of test, 
another for measuring the amount of sadness or depression expressed 
by the subject, another for classifying the nuances of tone of voice. 
None of these scales need any relationship to each other, and they 
leave unsolved the great paradox that personality must be considered 
as somewhat unified yet is expressed in a variety of ways. 

Actually, a distressing amount of creative energy has gone into 
molecular, stimulus-bound research of this sort. One method of clas- 
sifying the responses to one test, the Rorschach ink blots, involves over 
sixty elaborate and tricky rating procedures. These variables have 
direct reference only to the ink blots themselves, and by circuitous 
and generally unvalidated intuition refer to a few aspects of general 
behavior. This is a single example of the unfortunate and common 
practice of chasing one aspect of raw personality data down a tortuous 
side alley. 

We have several score of personality tests, each of which employs 
tedious methods for summarizing an extremely artificial and narrow 
range of expressive behavior. Most of these tests force the develop- 
ment of miniature personality theories which work for the tiny seg- 
ment of behavior that they tap. A test which uses sand and water as 
part of the stimulus items thus employs a theory which gives sand and 
water a prominent role in personality development. 

The solution we have employed to deal with this unsatisfactory 
situation is to define as the basic data of personahty, not the expressive 
events, but the communications by the subject or by others about his 
interpersonal activity. The basic units of personality come from the 
protocol language by which the subject's interpersonal behavior is 

When the subject smiles we attend to it, but the smile is not the 
datum which directly concerns us. Someone who is present in the 
situation, or observes it in cinematic form, has to make a protocol 


statement about this movement of facial muscles before it becomes a 
datum of personality. We study not the actual behavior, but the 
language about it (including the subject's language about it). 

This may sound, at first impression, like a restricting definition. But 
when we remember that we can obtain many descriptions of the same 
momentary event, it actually provides a systematic way of multiplying 
our knowledge. The smile, for example, might elicit many data sen- 
tences. The subject himself might describe his motive purpose at the 
moment as friendliness. The consensual report of many judges might 
agree in attributing friendly purpose to the smile. A suspicious relative, 
however, might judge it as smug or patronizing. A dependent relative 
might attribute tender sympathy. Thus, this facial gesture produces 
many protocol statements which provide interpersonal information 
about the subject's description of self and his social stimulus value to 

The basic data of personality studied by the interpersonal system 
are the verbal protocol statements about interpersonal behavior, 
i.e., the language in which the subject or others describe his inter- 
personal interactions, perceptions, and symbols. The diverse molecular 
responses — tears, bodily movements, test reactions — are the raw ma- 
terials. From them we obtain the building blocks for the scientific 
study of personality. These are units of classification — terms such as 
depressed, angry, confident. 

The Structure of Scientific Language 

In the methodological aspects of the science, we use a wide variety 
of empirical techniques to obtam the raw data of personality. We 
utilize these direct observations by converting them into systematic 
protocol language. Scientific study of personality consists in a study 
of the systematic language by which we describe the many facets of 
behavior. These conceptual operations refer to the formal aspects of 
the science. 

This important division of scientific procedures into empirical and 
formal propositions has developed out of the scientific philosophy of 
the twentieth century. Bertrand Russell and the Logical Positivists 
(Wittgenstein, Carnap, etc.) have helped to make the distinction be- 
tween the synthetic operational language, which refers to measurable 
events in the physical-social world, and the formal analytic procedures 
by which the language of science is organized. 

These two distinct types of scientific communications were rede- 
fined and a third pragmatic function added by C. W. Morris. This 
American philosopher claimed that all scientific activity can be studied 
as forms of the language of science. The general science which studies 


the entire field of scientific communication he calls semiotic. He 
defines three different functions of scientific behavior: ( 1 ) Semantics 
studies the relation of signs to objects and thus covers the empirical, 
experimental, and methodological aspects of science. (2) Syntactics 
is concerned with the relation of signs to signs, and involves the formal 
procedures of logic, syntax, and mathematics. (3) Pragmatics deals 
with the relation of signs to the users of signs. This branch of semiotic 
studies the functional and applied meaning of communicative behavior. 
Let us examine Morris' three functions in more detail. 

(1) Every science has unique methods and variables for dealing 
with its specific data. These variables and their relationships are de- 
scribed in terms of language. Thus, despite the great variations in what 
scientists do with their various data, the net result always involves 
communication or sign behavior. 

Certain general rules hold for all empirical investigations. Among 
these we include the need for unambiguous operational definitions of 
terms, and the need for public and repeatable measurements, pro- 
cedures, and the like. Morris calls these semantic rules since they gov- 
ern the relationship of signs to the empirical events. All sciences differ, 
but all must conform to the same standards of objectivity. 

(2) These empirical propositions which are related to observable 
and testable facts are crucially different from the formal prepositional 
structures of a science. The latter comprise systems which regulate 
the relationship of signs or language units. They have no empirical 
reference. Such formal devices are indispensable because they deter- 
mine how the researcher organizes his factual language. Mathematics 
and the logical deductive systems employed by modem science do not 
depend upon empirical proof. They are, in this sense, complex sets of 
terms which are inflexibly related to each other according to pre- 
established, assumed rules. The arithmetical statement "two times 
five equals ten," for example, is a predetermined relationship based on 
our original definition of what each of the terms means. This sentence 
is therefore empty of factual meaning. The psychoanalytic statement 
"the ego wards off instinctual impulses," is similarly formal, depend- 
ing on the assumed relationship of ego and instinct. It has no empirical 

(3) The pragmatic aspects of the language of personality delimit 
a broad and ramified field. They refer to the sociology of our psy- 
chological knowledge, its pohtics, its practical application in diagnosis 
and therapy. We have found it necessary to narrow the scope of the 
pragmatics of our system to the predictive function in the psychiatric 
clinic. We have selected the interpersonal framework because it ap- 
pears to be the most functional in terms of survival of the individual 



and a critical prediction of clinical events. In due course we shall 
attempt to show that every variable and every diagnostic category 
presented in this book has been chosen to predict directly the 
crucial aspects of the subject's future behavior — particularly with the 
future therapist. Thus we equate the pragmatics of personality psy- 
chology with prediction. From the standpoint of psychiatric opera- 
tions — the orientation of this book — nothing is so important as to 
have probability knowledge of the patient's future pattern of inter- 
personal behavior. This interpretation of the pragmatics of person- 
ality is, of course, the narrow sector of the broad field outlined by 
Morris that is most pertinent to a clinical psychology. 

With this threefold classification in mind, let us return to the dis- 
tinction between empirical and formal propositions. Since empirical 
statements are related to and are limited to observable events, and since 
formal statements are not, it is of critical importance to distinguish 
between the two types of propositions. Failure to do so leads to dan- 
gerous fallacies. These generally involve tautological formal state- 
ments which appear to be empirical assertions. The psychoanalytic 
phrase just quoted, for example, refers only to Freud's logical struc- 
ture of personality. It refers to the relationship between the language 
forms "ego" and "id" employed by Freud. The psychoanalytic lin- 
guistic system, which is the most ambitious yet developed in the field 
of personality, has restricted empirical reference. Those who employ 
Freud's verbal conventions often imply that they are making factual 
statements rather than logical tautologies. Cripphng confusions and 
meaningless communications will inevitably result if empirical and 
formal statements and pragmatic operations are not kept clearly 

If they are kept distinct several benefits accrue. The most im- 
portant of these is the general ordering of scientific activity. From the 
chaotic complexity of personality data emerge three broad and dis- 
tinct sets of operations — the empirical-methodological, the formal- 
logical, and the practical applications. Personality study currently 
faces these three challenging tasks: to measure objectively and mean- 
ingfully, to relate the obtained variables systematically and logically, 
and to apply the resulting knowledge with known predictive accuracy. 
We shall accomplish these objectives most efficiently by working 
within the principles of contemporary unified science. The rules for 
empirical methods (reviewed in Chapter 4) will guide our approaches 
to the raw datum, and its conversion into rehable language units. The 
formal principles will assist us in organizing our linguistic units. The 
goal of pragmatic applicability will encourage us to relate our sys- 
tematic knowledge to external events and to functional issues. Seen 


in this light, personality psychology becomes part of a unified general 

The purpose and outline of this book can now be restated in terms 
of these three categories. The remainder of this chapter presents 
some basic principles, some of which deal with a Logic of Personality. 
The two subsequent chapters survey the empirical and functional 
aspects of the field of personality. Chapter 6 and the two subsequent 
sections (Part II and Part III) return to the same issues, presenting 
objective methods for measuring interpersonal variables and formal 
notational systems for relating them. In parts IV and V the prag- 
matic themes assume priority as we apply the conceptual system to 
problems of interpersonal diagnosis in and out of the psychiatric 

The Selection of Personality Variables 

We began by noting the complexity of personality. From the 
philosophy of science we obtained three categories of scientific dis- 
course which help bring preliminary order to this diversity. This 
chapter goes on to present five working principles, which further 
assist in clarifying and systematizing the chaotic, fluid intricacy of 
human behavior. 

The first issue concerns the variables, elements, or conceptual units 
to be employed in dealing with the enormously diverse range of 
protocol sentences which describe interpersonal behavior. Every 
personality theorist has faced the formal questions of how many ele- 
ments or variables of personality are to be employed and how they 
are related. The first impression one might receive from many pre- 
vious theorists is that personality structure is very uncomplicated. 
Scores of dichotomous variables have been offered as the basic dimen- 
sion of human behavior — schizothymic versus cyclothymic, intro- 
verted versus extroverted, etc. As many three-way classifications have 
been popularized — lean, fat, muscular; intropunitive, extropunitive, 
impunitive; and the like. Most of these narrow conceptual solutions 
have quickly collapsed when asked to carry the heavy load of human 
variety. A broad collection of variables is a necessary answer to the 
question of "how many?" 

Another, more elaborate but ineffective, solution to the problem 
of basic elements is to employ one extremely broad, vague variable 
such as libidinal force or drive-towards-groivth. Motive concepts of 
this sort allow plenty of room for diversity but give no specific 
assistance to the empirical worker. 

A broad set of simple and specific elements (that we have here held 
to be necessary) leads to another formal requirement. Several such 


systems of variables have been developed by personality theorists. 
Many of the variables in these systems have tended to overiap each 
other, to overweight certain interpersonal behaviors, and to miss 
others. They have not been related to each other in a systematic 
order (i.e., on a continuum or scale). Henry Murray published (I), 
in 1938, an extensive list of human "needs" which has merited the 
considerable usage it has received. In a later publication, Murray has 
criticized his own eclectic collection of motive variables by proposing 
that social scientists "devote themselves more resolutely than they 
have so far, to the building of a comprehensive system of concepts 
which are defined not only operationally but in relation to each 
other." (2, p. 200) This demanding proposal, which we herewith 
include in our list of working principles, means that all variables 
should be related to each other along some kind of continuum. It 
means that each element should be located in fixed relationship to all 

Collecting the strands we have been weaving so far in this book — 
interpersonal orientation, adjustment-maladjustment continuum, sim- 
plicity, specificity, systematic relatedness — we are ready to state an- 
other working principle which guides our approach to human 

Third working principle: Measurement of interpersonal behavior 
requires a broad collection of simple, specific variables which are sys- 
tematically related to each other, and which are applicable to the 
study of adjustive or maladjustive responses. 

The Logic of Interaction 

Another formal issue must now be met. Interpersonal behavior has 
been defined as the basic area of personahty. It is in the essence of 
interpersonal phenomena that they never exist in isolation, but always 
in interaction with real or imagined others. We must conceive the 
interpersonal activity of the subject as he sees it, expresses it, and 
symbolizes it. We must, in addition, include his perceptions and sym- 
bolic views of others, as well as the responses which he pulls or obtains 
from others. An interaction psychology which deals with the issues 
of what-people-do-to-each-other runs headlong into another nest of 
classic philosophic entanglements — the subject-object dichotomy. 
Here we need another principle to clarify important issues. 

Fourth working principle: The interpersonal theory of per- 
sonality logically requires that, for each variable or variable system by 
which we measure the subject's behavior {at all levels of personality), 
we must include an equivalent set for measuring the behavior of each 
specified ''other'' with whom the subject interacts. 


In interpersonal psychology the simplest proposition is a two-way 
proposition. The subject is always in observed, perceived, or imagined 
interaction with crucial "others." These "others" may or may not be 
real persons. Considerations of methodological economy always limit 
the number and extent of the interactions that we can study. There- 
fore, some "others" never get measured or placed on the summary 

The Multilevel Nature of Personality 

We are engaged in this chapter in stating some working principles 
on which we shall base an adequate codification system for personality. 
The task of organizing personality data into logical categories reaches 
its climax when we face the problem of levels. 

Recognition of the multidimensional aspect of human nature is a 
landmark in the development of personality theory. Freud's demon- 
stration of the importance of unconscious motivation was an epochal 
intellectual achievement. The single-minded view of man as a rational 
being was supplanted by a binocular or multiocular vision of human 
character. It has revolutionized our concepts of personality. It has 
demonstrated that human behavior is not a unified single process; it 
is not just what it appears on the surface, nor what it is consciously 
assumed by the actor to be. It is rather a shifting, conflicted, multi- 
faceted complex of motives, overt and covert. 

The essence of modem personality psychology is its multidimen- 
sional character. 

Commonsense notions about human nature tend to be unilevel. 
People tend to think that what they consciously believe and say about 
themselves is the entirety of their personality. They are often quite 
unaware of intense and pressing emotions which dominate and direct 
their behavior. 

Experimental and academic psychology were untU recently com- 
pletely unilevel. The notion that what a subject reports is based on 
assumptions and motives which are not publicly stated came as a 
great surprise to the Behaviorists. 

Most of the current research in the field of personality is still dis- 
tressingly unilevel in its conception and research design. The standard 
instruments of personality research, the rating scale, the check lists and 
the Q-sort, can be rendered quite ambiguous by the introduction of 
multilevel logic. A typical research technique is to present a psy- 
chological judge with a test protocol — let us say an MMPI profile or a 
Rorschach record — and to ask him to rate the patient on a list of 
variables, or to sort a list of descriptive phrases about the patient. 
Multilevel logic requires that this task be rejected as meaningless. The 



questions are immediately raised: Should I rate how I predict he will 
behave, or how he will consciously see himself to be, or what I predict 
his underlying motives are? The simple, old-fashioned procedure of 
rating the subject thus breaks down into three or four rating ap- 
proaches, each of which may differ dramatically from the others at 
different levels. 

Many generalizations about results in personality research are 
similarly crippled by a unilevel approach. This is particularly true in 
the case of psychiatric and psychosomatic studies. Statements to the 
effect that obese patients are dependent, neurodermatitis patients are 
guilty and ulcer patients are passive, are quite limited in meaning. 
They seem to disregard the essential and basic concept of modern per- 
sonality theory — that the human being is a complex, multilevel pattern 
of conflicting motives and behaviors. The importance of a multilevel 
approach to personality can now be formalized. 

Fifth working principle: Any statement about personality must 
indicate the level of personality to which it refers. 

This is the key concept upon which this book is based. It will be 
noted in the clinical and descriptive sections of this book that no refer- 
ence is made to behavior without the accompanying designation of 
the level from which it comes. Thus we say that ulcer patients are 
responsible and managerial — at the level of overt public behavior; that 
hypertensive patients are sweet and affiliative — in their conscious self- 
description; that dermatosis patients are masochistic — at the level of 
imaginative fantasy; etc. 

The prudish (and often painful) circumlocution which this prin- 
ciple requires leads to a less graceful prose. It often puzzles and 
irritates the listener, who hopes to hear more definite statements about 
patients. In this connection we recall the staff meeting in which a 
psychosomatic research was being reported. An interested internist 
pressed for straightforward answers to his questions. "Are these 
patients passive and dependent?" The reply had to be cumbersome: 
"They are not at all passive at the two overt levels; they are sig- 
nificantly passive and dependent at the level of preconscious fantasy." 

Diagnostic language in the same fashion becomes multiplied in 
complexity when a multilevel approach is employed. We no longer 
find it possible to rattle off a single diagnostic label. To the question, 
"Is this patient schizoid?" a diagnostician using the interpersonal sys- 
tem of personality would respond in three-layer terminology. A 
typical answer might be: "At the level of symptomatic behavior the 
patient is phobic; at the level of conscious self-description, hysteric; at 
the level of the preconscious, intensely schizoid." 


We have discovered that it takes considerable patience and effort 
for psychologists to train themselves to think in multilevel terms. 
The behavioristic background of academic psychology apparently 
makes unilevel conceptions more congenial. Psychoanalysts, on the 
other hand, work comfortably and naturally in a multilevel idiom, 
although they are somewhat uneasy when their freedom to swoop 
from level to level is threatened by the limitations of operational 

The Logic of Levels 

The concept of multilevel behavior has immeasurably deepened our 
understanding of human nature. In addition to revising most of our 
psychological notions, it has broadened our interpretations of artistic, 
literary, and historical activity. Along with these intellectual boons, 
however, came a host of new problems and confusions. Much fal- 
lacious thinking has based itself on the conscious-unconscious dichot- 
omy. Formal systems for clarifying the illogical language of dynamic 
psychiatry seem to be needed. The next few decades will undoubtedly 
witness the introduction of many new systems of personality. Al- 
though the content of the theories may vary, it is hard to conceive of a 
personality theory (in this post-Freudian era) which does not deal 
with the problem of levels. It seems inevitable that systematic and 
logical rules must be developed for dealing with the multidimensional 
aspects of personaHty data. The following principles seem to be so 
axiomatic as to hold for all such personality theories. 

Sixth working principle: The levels of personality employed in 
any theoretical system must be specifically listed and defined. Once 
the logical system of levels and relationships among levels is defined, it 
cannot be changed without revising all previous references to levels. 

Illogical procedures will nullify the most brilliant concepts. Good 
logic, on the contrary, is one of the most powerful instruments we can 
use in forging a theory. The postulates just suggested for dealing with 
the problems of levels inevitably force an increase in theoretical pre- 
cision and scope. Listing and defining levels leads to improvements in 
empirical operations by clarifying the different sources of data con- 
tributing to each level of personality. This procedure has led us, for 
example, to the discovery that different probability laws hold for the 
different levels. Defining the formal relationships among the levels 
immediately reveals overlaps, tautologies, and previously undefined 
relationships of considerable theoretical promise. The conceptual 
issues of conflict, discrepancy, and motivating forces become sharp- 
ened. New conceptual entities become apparent. New research 


hypotheses develop. Indicating and consistently maintaining the levels 
of the data allow language usage to become more public and precise. 
A final and perhaps most important advantage of notational systems 
is that good logic breeds better logic. Any formal system should re- 
veal its own limitations and restricting assumptions. This, in turn, 
helps to father new and improved generations of successors. 

Multilevel Relatedness of Variables 

This chapter has been concerned with organizing the complexity 
of behavior into orderly classifications. Four working principles have 
been presented. They refer to variable systems and the levels of be- 
havior at which the systems are employed. Before this discussion is 
concluded, one final principle must be discussed. 

Seventh working principle: The same variable system should be 
employed to measure interpersonal behavior at all levels of personality. 

This means that we shall use the same classificatory elements regard- 
less of the level of the data. Most dynamic or multilevel systems of 
personality do not follow this suggestion. They employ one classi- 
ficatory language for covert, underlying themes and another language 
for describing overt behavior. 

There is a significant advantage in using the same variable system at 
all levels. It is possible to make direct comparisons between levels. It 
is possible to measure discrepancies, conflicts, or concordances among 
levels. These measurable indices of discrepancy, which we call indices 
of variability (some of which are like the traditional defense mech- 
anisms), are useful in several ways. They fill out our clinical picture 
of the personality by providing quantitative indices of the amount and 
kind of interlevel conflict. They are valuable indications of the inter- 
level organization of personality. They make possible objective re- 
search into such concepts as identification, repression, and idealization. 


The themes of this chapter are the complexity of personality and 
the requirements for dealing with it systematically. The general 
strategy to be employed should now be clear. First, we set up a broad 
variable system of interrelated variables. We use this to classify the 
interpersonal behavior of the subject and his world at several levels 
of personality. 

The essence of this approach is that we obtain thousands of single, 
specific, reliable molecular measurements. This makes for an objective 
system. We get at the complexity of personality by setting up the 


system of levels, then studying and comparing a pattern of hundreds 
of scores at the different levels. 

We do not employ clinical rating or intuitive judgments; although 
these are often broad, penetrating, and give a well-rounded picture of 
the personality, they are notoriously unreliable and unduplicable. 
For this reason we do not use professional psychological ratings at 
any point in the organizing of data. The procedure of automatically 
sorting thousands of reliable unilevel ratings into a standardized multi- 
level system allows us to pay some respect to the complexity of per- 
sonality without sacrifice of objectivity. 


1. Murray, H. A. Explorations in personality: A clinical and experimental study of 
fifty men of college age. New York: Oxford, 1938. 

2. Murray, H. A. Research planning: A few proposals. In S. S. Sargent and Marian 
W. Smith (eds.). Culture and personality. New York: Viking Fund, 1949. 

Empirical Principles in 
Personality Research 

In the preceding chapter it was asserted that logical procedures are 
required to order the data of any science. Prior to these formal opera- 
tions, however, comes the issue of collecting the data. This includes 
observing the raw events and performing some kind of discrimination 
or measurement. Empirical rules are required for this aspect of scien- 
tific activity. The interpersonal system of personality has attempted 
to follow three commonly accepted rules of scientific activity which 
can be formalized in a general working principle. 

Eighth working principle: Measurements of interpersonal be- 
havior Tnust be public and verifiable operations; the variables must be 
capable of operational definition. Our conclusions about human 
nature cannot be presented as absolute facts but as probability state- 

Personality Variables Must Be Public and Verifiable 

The first criterion of scientific activity insists that it must be public 
and verifiable. Any statement we make about the world of events 
must be subject to independent check. Its validity eventually rests on 
its confirmation by other scientists. While this social criterion of 
knowledge has engendered some qualifying controversy in the phi- 
losophy of the physical sciences, its employment in personality psy- 
chology at the present time is particularly necessary. 

Psychology, more than any other modem discipline, has been 
hampered by the issue of "private" observation. Many respectable 
scholars have flatly rejected the public testability principle and have 
endorsed a discipline of introspection, intuition, and anarchic indi- 
viduality. Many brilliant clinicians still stick by the principle that the 
human being is a unique and rather sacred pattern of individuality 



and that any attempt to find lawful generality is futile, insulting, and 
vaguely inhuman. 

The patient-oriented approach of the practitioner is highly credit- 
able, and needs no defense. To the clinician, the only principle in- 
volved is the welfare of the patient. There is, however, another 
important aspect to this question. Our technical competence to serve 
a patient is limited to our generalized, probabilistic lawful knowledge 
of human nature. Good will and patient-oriented solicitousness are 
virtues, but they are not professional instruments. Many skillful 
clinicians overlook the fact that they carry around inside of themselves 
a complex set of unverbalized and often unconscious generalizations 
about human behavior, which they apply to cases. Their patients get 
the benefit of an unsystematized lawful wisdom. These principles are 
often uncommunicable, unorganized, unreachable, untestable. They 
produce nothing toward the broad social goal of a science of human 

The integrity and productivity of good clinicians, however, more 
than justifies their unilateral approach at this primitive stage in the 
field. They violate no scientific canons because they do not pose as 

As soon as a clinician begins to lecture or write about principles of 
personality, however, he puts himself into the area of discourse that 
must be bound by the laws of scientific evidence. The first of these 
necessary conventions is that the events, the data, be open for inde- 
pendent verification by other scientists. 

There is a necessary objection which holds that psychotherapy can- 
not be studied objectively because the crucial events — the interpreta- 
tion, the instant resistance of the patient, etc. — cannot be repeated. 
This comment is quite beside the point. The data of personality are 
communications about human behavior — descriptions of the subject 
by himself and by others. The reliability and verifiability of these can 
be established by means of the most basic recording or data-preserving 
devices. The attempt to derive generalizations about human-person- 
ality-in-therapy probably will involve the use of objective electric 
recordings of the therapy process. 

With simple devices of this sort, it is possible to have any number 
of independent experts repeat and verify the most complex variable 
measurements. Without them psychotherapy becomes a wise but un- 
communicable art. WTien it becomes clear that the unit of per- 
sonality or interaction is the discriminatory element or variable, it also 
becomes quite feasible to obtain any number of equivalent repetitions 
of the variable by increasing the sample of subjects or of future obser- 
vations. While it is true that any raw personality expression is unique 



and unrepeatable, the basic variable units by means of which we clas- 
sify behavior are, by definition, general, recurrent, and verifiable. 

Operational Definitions of Terms 

A second and related aspect of scientific method which holds for 
personality psychology is that of operationism. This principle requires 
that terms be defined by the empirical operations which produce 
them. In the words of Bridgman, "We mean by any concept nothing 
more than a set of operations." The relationship between the terms 
we use and the empirical operations by which we discriminate them 
must be direct and openly expressed. 

In philosophy, the healthy impact of the operational definition has 
been to sweep away many metaphysical pseudoempirical concepts for 
which no external reference existed. In psychology many terms which 
have had dubious speculative histories have taken on new objective 
significance as researchers have linked their meaning to the empirical 
procedures by which they were measured. In personality and psycho- 
analytic theory — fields where undefined or privately defined concepts 
flourish like jungle growth — much less operational redefinition has 

There can be many operational definitions of the same concept. 
Each scientist may find it necessary to use different sets of data to 
define, for example, unconsciousness. One may use dreams. Another 
may employ fantasy stories, and another, slips of the tongue. As long 
as each worker clearly states the classificatory operations to which he 
relates his term there is no objection to the individual differences in 
approach. The rest of his colleagues are free to accept or reject his 
theories, but they cannot deny the empirical adequacy of his approach. 

Now, this flexibility of the definition process is not cause for alarm, 
nor is it a sign of any peculiar looseness of the personality field. The 
vahdity and meaning of any scientific fact is never exact or final. It 
always depends, among other things, on the type and level of the 
measurement methods involved. Only metaphysics can claim the 
luxury of finality and complete unambiguity. As the philosophers of 
operationism have pointed out, there are many ways to measure dis- 
tance — a yardstick, a mileage indicator, a transit reading. Each of 
these can be valid in its own area of discourse. Many of them can be 
combined into the same classification. Many cannot, at this point. 
Similarly our illustrative operational definitions of unconsciousness are 
(to the extent that they are independently confirmed) all valid. Many 
of them may be combined. It might, perhaps, be determined that 
dreams and fantasy stories tap the same level of unconsciousness, and 
allow a broader combined definition of unconsciousness. Slips of the 


tongue, possibly, might not be so related, and therefore would define 
another level of unconsciousness with its own particular lawful pre- 

The concepts of operationism have added powerful synthetic tools 
to the scientific method. Operational definitions have a remarkable 
capacity for ridding the language of any discipline of broad, impres- 
sive, but empty, terms which have no empirical meaning. Applied to 
the terminology of psychiatry, operationism calls for the elimination 
or systematic redefinition of almost every current concept. Operation- 
ism's "radical implications for psychiatric theory and practice" have 
been programmatically cited by MuUahy. He believes that "there is 
no chance that psychiatry will ever be a truly scientific field of inquiry 
until, as a first step towards scientific progress, it adopts a language 
sufficiently precise that its practitioners as well as workers in allied 
and related fields can in various ways check and verify the correct- 
ness of statements made by one another." ( 1, p. 58) 

The Probability Nature of Predictive Accuracy 

There is a third empirical principle which has importance for per- 
sonality psychology. This has to do with the ultimate validity of em- 
pirical knowledge. It holds that there is no absolute or final truth, 
that scientific laws are never completely accurate, and that the only 
knowledge we can have of the empirical world is probable knowledge. 
The essence of scientific explanation is the known relative accuracy 
of predictions. 

We tread here on the most ancient and hallowed ground of West- 
em philosophy — epistemological questions about the validity of 
knowledge. Within the last century statistical mathematics, post- 
Newtonian physics, and the operational logicians have produced con- 
verging solutions that are closely related to the needs and complexities 
of a functionalistic personality psychology. 

The most accurate statement any scientist can make about the 
world of events is an indication of the probability of occurrence. 
The chances are, let us say, three to five that a certain patient will 
develop passive resistance to a male therapist. But the chances are 
also two to five that he will not. Or we might determine that two 
thirds of the patients with duodenal ulcers will deny feelings of 
passivity and weakness. Of the one third who do not, 80 per cent 
manifest another specific interpersonal behavior — most likely schizoid 
withdrawal. When we have accumulated thousands of probability 
figures of this sort, based on publicly managed variable systems and 
organized into multilevel conceptual systems, a scientific structure of 
personality facts will have been established. Predictive procedures of 



limited but known accuracy will be at hand. Moreover, the com- 
plexity and variety of human nature need never be threatened by the 
necessary oversimplifications of our predictive structures. There can 
be as many different systems as there are different dimensions of per- 
sonality or of facets to the interreacting environment. The system de- 
scribed in this book is one such conceptual apparatus. It is designed 
to make factual predictions about the interpersonal dimension of be- 
havior in the clinical situation. This is really a very narrow slice of 
the wide and varied expanse of human behavior. Other systems will 
continue to appear. New variables will be developed. Broader areas 
of human behavior will be encompassed and integrated. The essence 
of scientific activity is that new theories, new facts never push out the 
old. They add, they revise, they refine, they expand. 

Thus we shall in later chapters present operational definitions of 
several psychiatric and personality variables and probability state- 
ments about their application. But no note of finality will be sounded. 
Future theorists will unquestionably present different and more effec- 
tive definitions of the same concepts — based on different operations 
and boasting, perhaps, higher probability relationships to functional 
criteria. To the extent that these varying approaches are objective — 
communicable and operationally grounded — the new findings will not 
disprove nor quarrel with the old. No scientific fact can be disproved. 
It can be reinterpreted, qualified by new relationships, amplified to fit 
new material. Scientific findings do not compete, debate, or attack 
each other. They add, expand, and collaborate to develop new 
hypotheses. This characteristic of the scientific method is particularly 
important in the study of human nature and has been often neglected. 


%. MuLLAHY, P. The theories of H. S. Sullivan. In P. Mullahy (ed.). The contribu- 
tions of Harry Stack Sullivan. New York: Hermitage House, 1952. 

Functional Theory of Personality 

The preceding four chapters have presented a sequence of principles 
which serve as background to a science of human nature. This chapter 
discusses the functional purpose of scientific knovi^ledge in general and 
psychological knowledge in particular. In so doing it calls upon and 
offers some synthesis of the principles already presented. There is 
more speculation and value orientation than in the preceding chapter. 

The Aims of General Science 

The ultimate objective of scientific activity is to explain and pre- 
dict. To control, change, cure, and improve are worthy motives. 
These latter tasks fall, however, within the province of the applied 
professions — engineering, administration, medicine, psychiatry. The 
job of the scientist is to explain as accurately and as completely as 
possible the relationships among variables and to predict future events. 

We explain any event by determining the probability relationships 
it has with other events. Increasing the temperature above a certain 
point is related to the boiling of water. Relationships of this sort in 
the macroscopic physical world have such regularity that extremely 
high predictability or exceptionless cause-effect sequences are gen- 
erally observed. The fields of atomic and subatomic physics and of 
human behavior involve such a multiplicity of interacting events that 
deterministic causal laws are not possible and probability statistics 
define the order of relationship. "The more rejecting the parents are, 
the higher expectation that the child will manifest a defensive sus- 
piciousness." Did the parents' rejection cause the child's distrust? It 
is much preferable to say that the two are correlated to a specific 

Probability laws allow us to make generalizations of known ac- 
curacy about the subject matter. Many established relationships among 
variables allow an increasingly higher order of generality. The breadth 
and sharpness of the explanatory process grow. 



But why do scientists attempt to explain natural and psychological 
events' What is the function of the generalized knowledge so ac- 
cumulated? These questions lead us to the other aim of scientific 
activity — prediction. 

The purpose of scientific explanation is to predict functionally 
useful events of the future. This conception of the scientist's role 
(which is, by the way, an opinion rather than an axiom) is a human- 
istic one. It assigns his social function in response to social demands 
and sees him as a human being always stimulated by and limited to 
cultural pressures. 

It is interesting to speculate that the human quest for knowledge 
has been strongly related to man's motivation to know the future. 
Knowledge of things to come has an enormous and obvious survival 
value. A major proportion of man's cognitive, philosophic activity is 
tied to his desire to anticipate correctly the future. Every religious 
interpretation has had to rest its dogma on a forecast about the nature 
of an afterlife. Much of its irrational and powerful appeal rests on 
this function. The interpersonal counterpart of these speculations 
might hold that ignorance is experienced as weakness, helplessness, and 
survivally dangerous. Knowledge is experienced as mastery and au- 
tonomy. It is survivally crucial in its function of forecasting the 

The time-bound essence of human life requires that man anticipate 
the things to come with reasonable accuracy. Science as the broad 
branch of human activity entrusted with the development and classi- 
fication of knowledge accepts the function of prediction. 

An activity often erroneously assigned to scientific activity is the 
function of control. Ideally there should be no reason why the ap- 
plication of pertinent knowledge to human problems should not be 
accomplished by the scientists who derive it. In actuality, the inter- 
personal behavior of human beings — particularly along the power axis 
— is so corruptible that there is good reason for the division of labor. 
Objective, effective scientific activity apparently suffers in direct pro- 
portion to the intensity of the interpersonal network involved. 

It is, thus, the task of the applied disciplines to use the predictive 
facts accruing from science. This distinction is not an invidious one. 
The years of technical training involved in the service professions — 
medicine, engineering — is often as great as or greater than that of the 
scientist. The responsibilities undertaken are invariably larger. So are 
the salaries. 

Neither is this distinction absolute. Most researchers employed by 
nonacademic institutions — whether industries or clinics — are generally 
forced to play a double role. They follow their scientific noses and 


are also led by them. This collaboration of the scientific with the 
applied is generally a fortunate one. Certainly for the problems of 
psychotherapy and personality change it is hard to see how much can 
be accomplished without complete clinical training as a minimum and 
considerable clinical practice as an optimum. 

Functional Theory of Personality 

To this point we have examined the functions of science in general. 
Turning to personality psychology we have seen the objective of this 
field to explain and predict interpersonal behavior. 

Objective empirical methods provide innumerable probability rela- 
tionships among specific variables. Formal and theoretical structures 
suggest how these are to be further related. This procedure poses new 
hypothetical questions. These are tested by additional empirical facts. 
This reciprocal progression of finding and theory establishes an in- 
creasing number of factual clusters which themselves become related 
to higher level theories. 

As understanding grows, the predictive power of the science be- 
comes more accurate and extensive. The functional importance of the 
field grows, usually encouraging new cycles of empirical activity. 

The complexity of human nature is such that there are countless 
facets of behavioral data and an equal number of empirical problems. 
The conceptualization and terminology of the field clearly depend on 
which of these aspects of personality are studied. The psychologist 
who spends all of his time measuring and relating variables of energy 
level will generally develop terms and theories that have something 
to do with energy. Even when we define personality in terms of the 
interpersonal behaviors, a broad scope remains. Every individual has 
been in crucial interaction with others since the day of his birth, and 
his history of past relationships is rich. Concentrating on the present 
rather than the past, we see an enormously extensive network of inter- 
personal reactions. Relationships in the family situation, in the job 
situation, or in the social sphere all have some explanatory value. In 
attempting to predict, which facet of social behavior should be focused 
on? We might be able to predict the interpersonal consequences of a 
subject's marriage to this girl, of his election to that office in the 
Masonic Lodge, or of the selection of a certain program of psycho- 
therapy in the clinic. The relevance of the prediction clearly refers 
to the problem being posed or the questions being asked. Prognostic 
knowledge is generally of value to the extent that it is relevant to the 
human problems at issue. To go further, it is most functional when the 
variables and terminologies of explanation are directly related to, or 


even in terms of, the functionally important activities. For clinical 
psychiatry this means that the variable language should refer most 
directly to the interpersonal interactions that determine a successful or 
unsuccessful clinical relationship. This point brings us to the question 
of functional diagnosis and deserves further illustration. 

Functional Diagnosis 

Let us suppose that a psychotherapist comes to the predictive diag- 
nostician posing this narrowly defined hypothetical problem. "In my 
office there is a male adult patient with asthma; what predictive state- 
ments can you make?" By studying the accumulated generalizations at 
hand the diagnostician might make any number of predictions. He 
might report, "The chances are better than two to one that your 
patient is married." This interpersonal prediction could be based on 
testable evidence, but it has little relevance to the situation at hand and 
little functional meaning. The diagnostician might report, "The 
chances are better than two to one that any asthmatic condition is 
related to psychogenic factors and is therefore psychosomatic." This 
is a descriptive, nosological statement. It has some relevance in that 
the psychiatrist can continue his clinical procedures with better than 
average chance that a psychological problem is related. It certainly 
does not throw much further specific light on the problem. 

A third possible answer might be, "Over 60 per cent of these 
patients during childhood show marked ambivalence toward the 
maternal figure and intense oedipal conflict with the father." This his- 
torical explanation is clearly more pertinent to the understanding of 
the patient. It might lead to extrapolating conjectures from the past to 
the future, and might assist in clarifying this patient's relationship to 
others including the future therapist. 

A fourth illustrative forecast might state that "Over 6$ per 
cent of asthmatic patients tend to be compulsively orderly and punc- 
tual." This is a testable psychological statement relating to the present, 
but it is molecular and peripheral, and has limited practical meaning. 
It is not directly interpersonal. It refers to stylistic symptoms rather 
than crucial purposive direction. 

None of these illustrative answers is adequately functional. They 
all can be true. They all might have some relationship to the per- 
sonality organization of the patient, but their bearing on the situation 
is not central. The pressure of the human problem at stake is not 
effectively met by these statements. In the clinical situation, a gen- 
eralized statement is most relevant to the extent that it predicts the 
future course of clinical progress. A diagnostic statement about a 



psychiatric patient is most functional to the extent that it forecasts 
interpersonal behavior pertinent to the therapeutic handling of his 

The patient cannot change his childhood experience, although it is 
very true that he can learn from it. The historical prediction is, thus, 
valuable, but not crucial. Nor is the patient's situation very dependent 
on diagnosing him psychosomatic. The diagnostic label is made by 
and is important to the clinician, and not to the patient. Changing this 
descriptive term would have very little effect on the symptom or the 
underlying character structure. Neither does the symptomatic molec- 
ular prediction about compulsive orderiiness have central importance. 
The punctuality and neatness are undoubtedly related to basic inter- 
personal motivations, but to focus on them diagnostically or thera- 
peutically would not be a recommended course of action. These 
stylistic "how" variables of personality take on their vital meaning 
when they are traced back to the interpersonal purposes which they 
serve. To change just the peripheral, noninterpersonal trait is not the 
essence of therapeutic improvement. 

The most functional answer to the clinician's question might go 
like this, "Over 75 per cent of male asthmatic patients who come to a 
psychiatric clinic manifest autonomous and stubborn competitiveness 
with males of superior or equal status. Conscious awareness of this 
intense fear of weakness is generally followed by overt signs of severe 
anxiety and increased competitive behavior. The chances are three to 
two that these patients will interrupt therapy in autonomous resist- 
ance." This prediction serves to illustrate the issues of relevant predic- 
tion and functional diagnosis.^ 

A statement of this sort is preferable for several reasons. It is inter- 
personal. It relates to the future; not just to one expected event, but to 
a sequence of interaction (which is related to a conflict between levels 
of personality). It relates the expected interpersonal pattern to an 
estimate of treatability. The diagnostic concepts are expressed directly 
in terms of predictive behavior which has bearing on the future treat- 
ment relationship. The future therapist is told specifically how the 
patient might be expected to react to the therapist and to the treatment 
process. His attention is directed to the interpersonal responses which 
have so much to do with the success or failure of the therapy plan. 

This last is an interesting sidelight of functional terminology. The 
predictive terms that a diagnostic system employs not only reflect its 
theoretical focus. They also exercise a subtle but marked effect on the 
subsequent use made of the information. If a theoretical system (and 

* See Appendix D for an illustration of a personality report employing the inter- 
personal system to make a practical prediction about a patient's behavior in the clinic. 



the diagnostic terms it sponsors) emphasizes past events of the case 
history, it is likely that the following discussions will tend to em- 
phasize these areas. If the predictions in the hypothetical case employ 
the language of compulsivity, punctuality, and the like, the facets of 
behavior may be unduly attended to in the interviews that follow. 
Suggestibility and selectivity of content cues are the constant errors of 
psychotherapy. The less experienced or the less flexible the therapist, 
the more influence accruing to this indoctrinating effect of diagnostic 

In clinical practice we assess the functional value of a personality 
or psychiatric variable in terms of the predictive value for facilitating 
the future clinical relationship. Terms which have high predictive 
value (even if indirect) tend to remain in popular use. Terms which 
have little predictive "cash value" tend to disappear. Every psy- 
chiatric term possesses a cluster of prognostic nuances which influence 
the intake and therapeutic diagnosis. Most of these predictive at- 
tributes are vague, unproven, often implicit, but they carry a stagger- 
ing load of responsibility. 

Schizophrenia, for example, brings to mind a host of prognostic 
associations, "not a good outpatient," "poor risk for brief therapy," 
"poor risk for psychoanalysis," "supportive or ego-strengthening 
methods favored," "long institutional treatment optimal," "generally 
slow prognosis," etc. These distillations of clinical wisdom are un- 
systematized, unverified probability statements about the future be- 
havior of schizophrenic patients. The original diagnosis is presumably 
based on other classes of variable cues. That is, the patient is originally 
diagnosed schizophrenic because of delusions, withdrawal, marked 
projections on or misperceptions of reality, and the like. Some psy- 
chiatrists hold that the best diagnostic sign indicating poor prognosis 
is the elicitation of hallucinatory material. 

This type of informal cUnical folklore is a necessary and healthy 
development in an infant field. The criteria of prognostic value (how- 
ever vague the variable relationships) indicate that the discipline is 
struggling toward a predictive status. As this process occurs the usage 
of certain terms with lesser prognostic power begins to diminish. They 
maintain only descriptive and administrative popularity. Hebephrenic 
is such a term. Outside of some crude differentiations from the folk- 
lore of the shock ward there is little prognostic specificity which dis- 
tinguishes this term from, let us say, catatonic. 

The most functionally important aspects of human behavior seem 
to be the interpersonal. To understand a human being is to have proba- 
bility evidence about his relationships with others (perceived, actual, 
or symbolic), about the durable interpersonal techniques by which he 



wards off anxiety, and about the reciprocal responses these techniques 
pull from others. To make meaningful predictions about a human 
being is to translate our explanatory data into statements as to the 
expected interpersonal behaviors in specific functional situations. 

Explanatory concepts which deal with instincts, body apertures, 
symptomatic manifestations, and peripheral stylistic traits have in- 
direct value to the extent that they can be related to interpersonal 
behaviors. It is not really of much use to a future therapist to predict 
that his patient will be punctual and not flick ashes on the rug. 

It seems quite possible that within a few decades the slowly 
evolving laws of pragmatic usage will establish interpersonal concepts 
as a popular and useful diagnostic language. Two possibilities suggest 
themselves here — the first is that direct interpersonal terms will replace 
the disorganized nosology of present-day psychiatry; the second is 
that the current terms will be redefined in interpersonal terms. If the 
first alternative is accepted, terms such as psychopathic personality or 
schizoid personality would disappear in favor of specific systematic 
interpersonal labels. According to the second alternative, psychopathic 
personality would be redefined operationally in terms of the rebel- 
lious aggressive criteria, and schizoid personality would have as its 
basic diagnostic indices distrust and bitter withdrawal. This is another 
historical issue that time will settle. 

The system described in this book employs the latter — more con- 
servative — solution for developing a functional, operationally defined 
language of personahty which will work for both adaptive adjustment 
and the psychiatric extremes. 

Functional Co?icept of Personality 

Two general postulates, which have been woven in as background 
for all of the discussions so far, hold that the functional core of human 
behavior is the interpersonal, and that personality concepts must be 
defined along adjustment continua which include both normal and 
abnormal reactions. When we approach the problem of a functional 
personality language with these two principles in mind, certain solu- 
tions seem to follow quite readily. 

The first assumption clearly demands that the basic set of personal- 
ity variables be not symptomatic, erotogenic, or stylistic, but inter- 
personal. The second assumption suggests that each of these variables 
must have an intensity dimension such that its rigid, maladaptive ex- 
treme be as readily classified as its moderate adaptive aspect. The 
measurement categories all along this scale are still interpersonal — as 
we recall from the hostility continuum described in Chapter 2, where 
blunt, frank, appropriately critical were terms referring to the adaptive 


and sadistic, aggressive to the maladaptive end of the continuum. Now 
it is well known that the language of psychiatry deals almost ex- 
clusively with the pathological extreme of behavior. Thus we dis- 
cover that maladaptive extremities of the continuum for each generic 
interpersonal motivation are most closely related to and overlap the 
psychiatric. In the illustration of the hostility continuum just men- 
tioned, it will be noted that sadistic and aggressive have a much more 
psychiatric flavor than do blunt, frank, appropriately critical, and 
the like. 

It seems to follow, then, that if we painstakingly study all the forms 
of interpersonal behavior in as many environmental situations as pos- 
sible, we shall obtain, after grouping and sifting, a finite number of 
discernible basic interpersonal motivations all of which must (ac- 
cording to the normality assumption) be placed on adaptive-mal- 
adaptive continua. For each pathological interpersonal pattern we ob- 
serve in the clinic there must be an adjustive aspect. And for each 
successful social maneuver we meet in the market place there must 
be a pathological extreme. The surprising linguistic imbalance which 
implies that an Anglo-Saxon cannot be too affectionate or adaptively 
disaffiliative has already been commented upon. The implications of 
this imbalance for systematic functional diagnosis will be developed in 
later chapters. 

Since the neurotic interpersonal intensities tend to overlap some 
aspects of the noninterpersonal psychiatric categories, we have close 
to hand a solution for the problem of what to do with these latter less 
functional terms. The process of redefining them begins to take place 
automatically. Most of the popular diagnostic labels have vague, un- 
defined, but fairly effective functional power. They have interpersonal 
correlates. To be skeptical, realistic, and reserved is generally an 
adaptive interpersonal pattern. To be inflexibly distrustful and with- 
drawn is invariably maladjustive. Many psychiatrists would call it 
schizoid. Thus we see the possibilities of redefining the classical 
language of administrative psychiatry in interpersonal terms. This 
preserves the usefulness of the older terminology while sharpening 
its denotive power. On the other hand, from the standpoint of the 
interpersonal system we have added a new set of partially interpersonal 
terms to our linguistic structure which is broadened thereby. The ex- 
treme points of the scales now have a new set of descriptive terms 
which are unique to the professional specialists of the clinic but which 
relate to the broader system of general interpersonal psychology. An 
interpersonal notational system holds the promise of bridging the an- 
cient and logically intolerable gap between the science of personality 
and the practice of psychiatry. 


There will probably be many such reciprocal rapprochements in 
the next phases in the study of human nature. The scientist or systema- 
tist will do well, we suggest, to keep his general concepts from being 
swallowed up by the more exciting linguistics of the clinic. It is most 
valuable to stress the relationship between general concepts of per- 
sonality and the terminology of the practitioner. It is important, how- 
ever, to maintain the basic nature of the generic interpersonal systems. 

If this is done, the possibilities of relating the general sciences of 
interpersonal behavior with other applied and pure disciplines in addi- 
tion to clinical psychiatry appear bright. There is, for example, good 
reason to feel that occupational adjustment is mainly determined by 
interpersonal factors. Whether the applied field is vocational counsel- 
ing or industrial management, the terminology of job classification is 
very likely to have interpersonal correlates, with, perhaps, even more 
overlap than psychiatric labels. These vocational "diagnostic" terms 
are most likely to be located near the adaptive and moderate end of 
the normality-abnormality continua — blunt, frank, realistic, amiable, 
etc. A similar cross-fertilization and functional application seems 
quite feasible. Wherever an applied discipline requires psychological 
(not physiological) answers to the problems it faces, an interpersonal 
psychology will generally be best equipped to make the most basic 
explanations and the most functional predictions. 

Thp Working Principle of Functional Applicability 

The functional orientation which has just been described can be 
summarized in the form of a guiding statement. 

Ninth working principle: The system of personality should be 
designed to measure behavior in the functional context {avhich in this 
book is the psychiatric clinic). Its language, variables, and diagnostic 
categories should relate directly to the behavior expressed or to the 
practical decisions to be made in this functional situation. The system 
should yield predictions about interpersonal behavior to be expected 
in the psychiatric clinic. 

General Survey of Interpersonal 
and Variability Systems 

The preceding five chapters have presented a general, theoretical 
discussion of some of the basic requirements of an adequate science of 

By way of summary the nine working principles which have guided 
the Kaiser Foundation research in personality will now be reviewed 
before surveying the personality system. 

Nme Working Principles for 

the Interpersonal Theory of Personality 

(1) Personality is the multilevel pattern of interpersonal responses 
(overt, conscious or private) expressed by the individual. Interpersonal 
behavior is aimed at reducing anxiety. All the social, emotional, inter- 
personal activities of an individual can be understood as attempts to avoid 
anxiety or to establish and maintain self-esteem. 

(2) The variables of a personality system should be designed to meas- 
ure—on the same continuum— the normal or "adjustive" aspects of behavior 
as well as abnormal or pathological extremes. 

( 3 ) Measurement of interpersonal behavior requires a broad collection 
of simple, specific variables which are systematically related to each other 
and which are applicable to the study of adjustive or maladjustive 

(4) For each variable or variable system by which we measure the 
subject's behavior (at all levels of personality) we must include an equiv- 
alent set for measuring the behavior of specified "others" with whom the 
subject interacts. 

(5) Any statement about personality must indicate the level of per- 
sonality to which it refers. 

(6) The levels of personality employed in any theoretical system must 
be specifically listed and defined. The formal relationships which exist 
among the levels must be outlined. Once the logical system of levels and 



relationships among levels is defined it cannot be changed without revising 
all previous references to levels. 

(7) The same variable system should be employed to measure inter- 
personal behavior at all levels of personality, 

(8) Our measurements of interpersonal behavior must be public and 
verifiable operations; the variables must be capable of operational defi- 
nition. Our conclusions about human nature cannot be presented as 
absolute facts but as probability statements. 

(9) The system of personality should be designed to measure behavior 
in a functional context (e.g., the psychiatric clinic). Its language, variables, 
and diagnostic categories should relate directly to the behavior expressed 
or to the practical decisions to be made in this functional situation. The 
system, when used as a clinical instrument, should yield predictions about 
interpersonal behavior to be expected in the psychiatric clinic (e.g., in 
future psychotherapy). 

In the next six chapters (which comprise the second section of the 
book) these postulates will be employed in an attempt to construct 
such a system. The nature of these requirements tends to determine 
and limit the resulting personality system. In this chapter the over-all 
organization of the personality system will be described in terms of 
(1) a schema for classifying interpersonal behavior and (2) a formal 
notational system for defining and relating the levels of personality. 
The subsequent chapters will focus respectively on five levels of per- 
sonality and the way in which they are combined and used for inter- 
personal diagnosis. 

Before presenting the outline of the personality system, let us illus- 
trate by way of review the importance of formal theory for dealing 
with the levels of personality. Some remarks by the philosopher 
Reichenbach (on the value of symbolic logic) may be appropriate in 
this connection. He suggests that: 

The introduction of a symbolic notation is important to logical procedure 
because "it has about the same significance as a good mathematical notation." 
Suppose you are given the problem: "If Peter were 5 years younger, he would 
be twice as old as Paul was when he was 6 years younger, and if Peter were 
9 years older, he would be thrice as old as Paul, if Paul were 4 years younger." 
Try to solve it in the head by adding and subtracting and considering all the 
"if's," and you will soon arrive at a sort of dizziness as though you were 
riding on a merry-go-round. Then take a pen and paper, call Peter's age x and 
Paul's age y, write down the resulting equations and solve them the way you 
learned it in high school— and you will know what a notational technique is 
good for. There are similar problems in logic. (10, p. 219) 

There are also similar problems in dynamic psychology. Consider 
this not atypical case report from a psychoanalytic journal. The 
author describes a multilevel pattern of the patient's emotions as fol- 


While expressing aggression toward a male cousin, she thought once again 
that she smelled gas. At first by allusion to others, then by way of dreams, 
there emerged the fantasy that the analyst was feminine; then she admitted 
never having thought of her father as a man, but as a woman. 

She wished she could dominate the analyst and others as she felt dominated 
at home. This aggressive urge was accompanied by increased feelings of guilt. 
At a time when she had unconscious conflicts about not paying for cancelled 
hours, and also had arranged for more advanced art lessons which would 
increase her abilities and prestige, she stuck two fingers into an electric fan, and 
was unable to work. 

Seductive fantasies toward the analyst, as well as homosexual dreams and 
fantasies, and dreams of being gassed and raped emerged in connection with 
memories of compulsive masturbation in her childhood, causing vaginal dis- 
charge which she had had impulses to eat. After confessing her "dirty thoughts" 
she had a dream. 

"She stood before a mirror admiring herself, dressed in a beautiful 
flowing white dress." 

She said this dress made her look "effeminate" and then felt embarrassed at 
the use of the word. She felt that to be beautiful would serve two purposes: to 
make her sister and other girls feel inferior to her, and to control men. She had 
often thought mouth and vagina were equivalent. 

After this dream she became cleaner, worked better, and began to earn her 
way both by art work and by working in a department store. Competitive 
strivings in regard to other patients, as well as her sister, came out more clearly 
in association to wishes to be dirty. (7, p. 79) 

If the reader attempts to organize this series of conflicting events, to 
sort out the levels and the motives which belong to them, ambivalent, 
autistic, past, present, he may acquire a sort of vertigo similar to that 
mentioned by Reichenbach. 

This analyst has combined at least four or five levels of behavior in 
this passage. He describes certain overt actions of the patient: "ex- 
pressing aggression," "arranged for art lessons," "stuck two fingers 
in a fan," "worked better," "began to earn her own way." All of 
these actions are public — and could be consensually validated by 
listeners or observers. 

The analyst also mentions certain wishes, urges, or impulses which 
the patient reported: "to dominate the analyst," "to eat," "competitive 
striving." These impulses, consciously recognized but not acted out, 
must be kept systematically distinct from the above-mentioned overt 

Another level at which this patient operates is that of dream or 
fantasy: "that the analyst was feminine," "her father as ... a 
woman," "seductive fantasies" toward the analyst, as well as homo- 
sexual dreams and fantasies, and "dreams of being gassed and raped," 
etc. These autistic productions are clearly deeper or further from real- 
ity than the overt activities or the secret wishes previously summarized. 


To these three levels we might also add the deeper unconscious 
conflicts and the conscious reports — both of which denote different 
orders of reality-contact and consciousness. 

Free association protocols, case histories, and reports of therapeutic 
interaction comprise important sources of data upon which the science 
of personality must be built. In order to make reliable measurements, 
valid judgments, and meaningful analyses the multilevel jumble of 
motivations which so often characterizes personality descriptions must 
be organized into a systematic language. 

In this chapter we will describe first a classificatory system for or- 
dering interpersonal behavior. Then we shall present a notational sys- 
tem — a crude mathematic or grammar of personality — which attempts 
to order the levels of behavior. We shall present the units or variables 
by which the behavior can be measured, and five levels at which they 

The classificatory system allows us to measure interpersonal be- 
havior at any of these five levels. The notational schema defines the 
levels and the fixed arithmetic relationships among these levels. It 
provides for the diagrammatic and numerical analysis of the personal- 
ity structure. 

The Classification System: The Interpersonal 
Variables of Personality 
In beginning the long task of developing a personality system, the 
first assumption refers to the kind of behavior to be studied. We have 
defined this as the interpersonal core of personality. The initial step 
for the Kaiser Foundation research project was, therefore, to focus on 
this dimension of behavior. To this end a wide assortment of raw in- 
terpersonal data was assembled. Several scores of individuals — male 
and female, neurotic, psychosomatic, and normal — were brought into 
interpersonal relationships in small groups. Some of these were dis- 
cussion groups in a nonpsychiatric setting. Some were psychotherapy 
groups in an outpatient clinic. The hundreds of interactions of each 
subject were observed, recorded, and studied. Many other types of 
interpersonal behavior were obtained from the same subjects. Their 
verbal descriptions of self and others — present, past, and anticipated 
— as expressed in the groups or as summarized in autobiographies and 
psychological inventories were collected. Their dreams and fantasies 
were recorded. Their responses on batteries of projective tests were 
elicited. A rich but unwieldy collection of raw materials — in the form 
of wire recording spools, typed transcriptions, ratings, observers' re- 
ports, test indices, projective responses — piled up for each subject. 
In line with our first theoretical assumption, the interpersonal aspects 


of the stimulus material were taken as the focus of attention. As the 
research team observed this undigested mass of protocol records ac- 
cumulating, the next research question occurred. How shall we 
analyze these data? It was clear that classiiicatory assistance was re- 
quired. This came in the form of the second working principle, which 
holds that the basic data of personality are not the raw responses but 
the units of protocol language by which the subject's interpersonal 
behavior can be summarized. 

The selection of this language, as we have seen, has been a crucial 
aspect of all personality theories. What and how many are the con- 
ceptual units of social interaction? The third working principle en- 
ters at this point, stating that measurement of interpersonal behavior 
requires a broad collection of simple, specific variables which are 
applicable to the study of adjustive and maladjustive responses. 

With these guiding principles in mind, the diverse data were studied 
to determine the optimal number of specific variables and their orderly 
relationship. As a first step the interactions of the subjects were studied 
by three independent judges who attempted a straightforward verbal 
description of the interpersonal activity. In rating the observed and 
recorded interactions, it was noticed that transitive verbs were the 
handiest words for describing what the subjects did to each other, 
e.g., insult, challenge, answer, help. In rating the content of the spoken 
or written descriptions of self-or-other, it was noted that adjectives 
were more often suitable. Here we were interested in the attributes, 
qualities, and traits which the subject assigned to himself and others. 
"I am friendly, helpful, strong; they are hostile, selfish, wise, helpful^ 
A clear relationship seemed to exist between these two types of inter- 
personal description, such that the adjectives seemed to express an 
interpersonal attribute or potentiality for action, while the verbs 
described the action directly. Three rather interesting notions began 
to develop out of this fact. First, the relationships between different 
expressions of personality can be directly related to each other by 
grammatical or linguistic procedures. That is, what you actually do 
in the social situation as described by a verb (e.g., help) can be re- 
lated to your description of yourself (as described by the attribute 
helpful) and to your description of your dream-self or fantasy-self 
(also attributive, helpful or perhaps unhelpful). These grammatical 
relationships became the key to a systematic consideration of the 
levels of personality, of which more later. 

After extensive informal surveys of the many varieties of data, a 
list of several hundred terms for describing interpersonal behavior was 
assembled. The next task was to sort through the long lists of terms 
and to determine the generic interpersonal motives. Combining the 


action verbs with the corresponding attributive adjectives cut down 
the list. Thus the adjective insulting was subsumed under its action 
category to insult. Next the intense and statistically rare terms were 
combined with the moderate and more frequent categories. For ex- 
ample, the themes of murder, attack, insult, etc., were included under 
the generic concept of hostile activities. The gradually developing 
lists of generic terms were then combined to eliminate overlaps and 
repetitions until a list of sixteen generic interpersonal motivations re- 
sulted. All of the original terms — which numbered several hundred — 
could be expressed as differentiated varieties of the sixteen basic inter- 
personal themes. In this manner the goal of breadth, specificity, and 
simplicity was approached. 

The principle of systematic relatedness then determined the next 
task. This criterion demands that the variables be ordered along 
continua in such a way that fixed relationships exist between the ele- 
ments. The question here becomes: What and how many are the di- 
mensions along which the variables are to be scaled? In this instance, 
it became apparent that a two-dimensional grid was optimal for re- 
lating the variables at hand. We cannot doubt that more complex for- 
mal systems will eventually add new spatial dimensions to the organi- 
zation of personality. For the present, however, a two-dimensional 
space offers sufficient complexity for the data and more than a; suf- 
ficient complexity of methodological problems. 

In surveying the list of more or less generic interpersonal trends, 
it became clear that they all had some reference to a power or affilia- 
tion factor. When dominance-submission was taken as the vertical axis 
and hostility-affection as the horizontal, all of the other generic inter- 
personal factors could be expressed as combinations of these four nodal 
points. The various types of nurturant behavior appeared to be 
blends of strong and affectionate orientations toward others. Dis- 
trustful behaviors seemed to blend hostility and weakness. Further 
experimentation and review of the raw data led to the conclusion that 
a circular two-dimensional continuum of sixteen generic variables rep- 
resented the optimal degree of refinement of interpersonal themes. 
Attempts at more specific systematization of interpersonal behavior 
by increasing the number of variables led to difficulties in establishing 
clear criteria for discrimination between neighboring variables. On 
the other hand, use of grosser units of discrimination, e.g., only the 
four nodal variables, resulted in neglect of important shadings of inter- 
personal intent. 

The sixteen generic interpersonal themes are presented in Figure 1 . 
Each one has been assigned a code letter. Thus, Dominant behavior 
is classified under the letter A, Autonomous behavior under the letter 


B, etc. Several suggestive terms are listed for each generic type of 
interpersonal purpose in Figure 1. Actually, there is an almost inex- 
haustible list of terms for each generic code letter. The many varieties 
of interpersonal behavior included under each category will become 
increasingly clear as we take up the ratings for the different levels. 

,#^ \VSoc 










^,^'^^ /,fti 

^\ iL^ •*>.>* 






/V i\^ 



■Oft ■>> 




IrX \^^ 

^PtJ^/v / PROVOKES ' 


^Cns ^Cr^'^'^OA, / SPINELESS \0^^o T\\>^1=. >^^ 
^Oe^O'Vs'^^/ ACTIONS, y'^^oOS^O^^^^V^ 

c^^Cj'" sA . / SUBMIT \ cO^><;^^' -J/ 



Figure 1. Classificaiion of Interpersonal Behavior into Sixteen Mechanisms or 
Reflexes. Each of the sixteen interpersonal variables is illustrated by sample behaviors 
The inner circle presents illustrations of adaptive reflexes, e g., for the variable A, 
manage. The center ring indicates the type of behavior that this interpersonal reflex 
tends to "pull" from the other one. Thus we see that the person who uses the reflex A 
tends to provoke others to obedience, etc. These findings involve two-way inter- 
personal phenomena (what the subject does and what the "Other" does back) and are 
therefore less reliable than the other interpersonal codes presented in this figure. The 
next circle illustrates extreme or rigid reflexes, e.g., dominates. The perimeter of the 
circle is divided into eight general categories employed in interpersonal diagnosis. 
Fach category has a moderate (adaptive) and an extreme (pathological) intensity, 
e.g., Managerial-Autocratic. 


By arranging a set of sixteen interpersonal variables along a con- 
tinuum, we have implied a systematic relationship among them. If 
we rate any behavior as C, we have defined it in terms of all the other 
variables since C is one unit away from (and therefore close to) D and 
B, while it is eight units (and therefore quite discrepant) from K. The 
second working principle, which requires a relatedness among var- 
iables, is thus met but it is next required to demonstrate that the hypo- 
thetical relationships of these variables is related to external events. 
Extensive vahdation of the circular continuum of sixteen interpersonal 
variables has demonstrated that it is satisfactorily congruent with em- 
pirical facts. (5) (8) While the units around the scale are not com- 
pletely equidistant, the arrangement is correctly ordered. 

The selection and formal organization of variables made it possible 
to rate any interpersonal behavior in such a way that its relationship to 
all the other fifteen variables was explicit. The classificatory schema 
at this stage of the game was still far from complete. Only the most 
crude appraisals of any interpersonal behavior could be made because 
only the presence or absence of the theme could be indicated. For 
example, it was possible to say that distrust was present; but how 
much, how extreme, how inappropriate could not be measured until 
an intensity dimension was added. 

In the most basic sense this involved making a "more or less than" 
judgment of the observed event. Is this behavior more distrustful than 
the other? The intensity dimension is quite fundamental to all human 
perceptions. Language and quantitative usages give us several tech- 
niques for expressing intensities: the comparative sequence stroni!;, 
stronger, strongest, the modifying function extremely, slightly, as well 
as the intensity hierarchy of different word meanings critical-angry- 
furious-enraged. Apart from these verbal expressions, the numerical 
estimation of intensity (along a 3-, 5-, or 7-point scale) is accepted and 
common. The intensity of interpersonal activity can be rated on a 
linear scale ranging from absence of the behavior to extreme over- 
reactivity. The number of differentiating points on the intensity scale 
can vary according to the specific purpose, but for most interpersonal 
responses, a 3- or 4-point graduation seems quite satisfactory. 

Let us consider, by way of illustration, one interpersonal motivation 
as it is reflected in the intensity dimension. The power continuum 
(variable A) is conceived of as a linear scale ranging from too much to 
complete and inappropriate absence of dominance. When we con- 
struct an intensity scale for each of the sixteen interpersonal variables, 
we obtain a more differentiated form of the circular continuum which 
is illustrated in the concentric rings of Figure 1. The term dominate 
now takes on quite a precise meaning. It is defined as an expression of 


power (A) which systematically relates it to the other fifteen inter- 
personal themes. It is further assigned an intensity loading which re- 
lates it to all other verbal terms for power as well as to every other 
classified word describing interpersonal interaction. In this way 
language of personality becomes much more exact and accessible. 
Every term in the English language which refers to interpersonal be- 
havior can, in this manner, be studied, redefined systematically, and 
calibrated. This is not to say that these terms as used in everyday 
life necessarily have the same meaning to the interpersonal scientist. 
The general public employs all kinds of words — force, power, effi- 
ciency, hostility — which have been operationally redefined by physi- 
cal or psychological scientists. The interpersonal diagnostician dealing 
with human communications has to keep clear the level of meaning of 
the words he deals with. Anger may denote one thing to an individual 
patient, another in terms of general usage, and a third in the precisely 
defined scientific discourse. In general, it seems best to keep the scien- 
tific meaning as close as possible to that of the general public of the 
culture being studied. The advantages of tying terminology to func- 
tional behavior rather than tying it to psychiatric usage have already 
been mentioned. 

The two-dimensional representation of interpersonal space has 
many possibilities for summarizing behavior. First, it should be noted 
that we are rarely interested in classifying single, isolated events. In- 
variably we are concerned with sequences of interaction and patterns 
of hundreds of interpersonal expressions. The simplest and perhaps 
least useful way of summarizing interpersonal behavior is to plot the 
ratings, judgments, or units directly onto the circle. Suppose we re- 
cord and then rate the interpersonal purpose involved in everything 
a patient does to his analyst in the first twenty hours of therapy. This 
would produce (depending on the consistency and expressiveness of 
the patient) between 1,000 and 3,000 interpersonal units. Disregard- 
ing the intensity ratings, we thus obtain the total of all Dominance {A) 
ratings and the comparable totals for the other fifteen interpersonal 
themes. By calibrating the sixteen radii for numerical frequency, we 
can then strike off points indicating the reactions for each inter- 
personal variable. A graphic summary of the interpersonal behavior 
during twenty hours of therapy is thus obtained. In Figure 2 we see 
that the sample patient manifested docile, cooperative dependence 
toward the therapist, avoiding hostility and competitiveness. Pro- 
files based on other patients or upon this patient's behavior in the sub- 
sequent hours of treatment would allow direct, objective comparisons 
and the testing of hypotheses about interpersonal activity during 



When enough cases have been studied to provide normative data, a 
second and highly profitable method of summarizing interpersonal be- 
havior is possible. There are many statistical techniques for treating 
each patient's scores in terms of the mean (i.e., the average) of his 
group. These allow us to determine one point which summarizes all 

Figure 2. Diagrammatic Representation of Interpersonal In- 
teraction of a Patient During Twenty Hours of Psychotherapy. 
Radius of circle equals 1,000 interactions. This patient manifested 
820 docile-dependent interpersonal actions {JK octant) and 260 
confident-narcissistic actions {BC octant). 

of the interpersonal behavior in any behavioral sequence in terms of 
its distance and direction from the center of the circle. The latter 
is taken as the mean, i.e., the central tendency of the interpersonal be- 
havior of the population studied. One method for obtaining this 
summary point has been described as follows: 

The Interpersonal System as described so far leaves us wide latitude with respect 
to the formal (algebraic) properties which are to be attributed to the 16 
variables. We may in fact vary the formal relationships to suit the particular 
context so long as we do not violate the rough intuitive specification of a 
circular arrangement. For example, we might think of the system as a purely 
ordinal array about which one specified only that categories adjacent to a given 
one resemble it more than do non-adjacent categories. Or we might consider 
the circle to be a two-dimensional array in ordinary Euclidian space, in which 
case conventional trigonometric and analytic formulas relate the 16 variables. 
After some experimentation, this latter approach was tentatively selected. Each 
circle was conceived to be a set of eight vectors or points in a two-dimensional 
space. We selected the center of gravit)^ or vector mean of these points as a 
measure of central tendency. 

A vector in two-dimensional space may be represented numerically by the 
magnitude of its components in two arbitrarily selected directions. We chose 
AP and LAI as reference directions, giving the designations Dom and Lov 


respectively to the components of the vector sum in these two directions. 
Representation of the eight or sixteen scores comprising a patient's circle by a 
single point in two-dimensional space is a considerable simplification. What is 
preserved in this simplification is the general tendency of the circle. What is 
lost are the individual fluctuations around the circle.^ 

The formulas for the two components of the vector sum are relatively 
evident. They are: 

1. Dom = i«Ri sin ^, and 

\= 1 

2. Lov = i«R, cos 0i 

i= 1 

where R, = the score in the 1-th category, 

01 = the angle made by moving in counter-clockwise direction from L 
to the i-th category (from LM if octant scores are used). 

In the present calculations, octant scores were used and .7 was taken as the 
value of sin 45°; the following simplified formulas resulted: 

3. Dom =AP -HI +.7 (NO + BC - FG - JK), 

4. Lov =LM-DE + .7 (NO - BC - FG + JK), 
where AP = score in octant AP, etc. (4, p. 140) 

It is thus possible to convert the pattern of scores on the sixteen 
variables into two numerical indices which locate a subject's inter- 
personal behavior on a diagnostic grid. Figure 3 presents the descrip- 
tive summary point for the therapy patient whose behavior has been 
previously diagramed in Figure 2. We note that the two summary 
indices place him in the JK octant; they thus become a simplified and 
numerical summary of the circular diagram. The vertical and hori- 
zontal lines represent varying discrepancies from the mean (the center 
point of the circle). We obtain in this manner a circular grid, every 
point on which is statistically defined. We determine the summary 
point of the patient's interpersonal behavior as rated by the sixteen 
variables in relationship to the population studied — which in this case 
might be a hundred randomly selected psychotherapy patients. Our 
subject is seen as considerably more trustful and compliant than the 
average therapy patient (point 1 in Figure 3). 

The great advantage of the latter circular grid method of summari- 
zation is that many summary points can be graphed on the same dia- 

' The two components of the vector sum must each be divided by N = Ri (the 

i= 1 
total around the circle all eight or sixteen scores) to get the two components of the 
vector mean. These latter may also be thought of as the first two Fourier coefficients 
of a curve fitted to the observed data. More complicated curves can be fitted by the 
computation of additional coefficients. 



gram, facilitating comparison among levels of any individual's per- 
sonality or comparisons among different individuals. Let us suppose 
that the psychotherapy patient we have been using for illustration 
shifted his interpersonal behavior markedly in the second twenty hours 


^(Hjr ' 

Figure 3. Diagram Summarizing the Interpersonal Behavior of 
Patient During First Twenty Hours (T) and the Second Twenty 
Hours ® of Psychotherapy. Summary points are located by inter- 
section of horizontal and vertical indices. The indices are deter- 
mined by the raw number of interactions converted to vector 
scores by the trigonometric formulas described on page 69. 

of treatment, expressing disappointment and distrust towards the 
analyst. The several thousand interactions are rated, statistically sum- 
marized, and graphed as point 2^on Figure 3. A diagrammatic con- 
densation of the changing behavior of the patient (based on quanti- 
tative objective methods) becomes available. This patient has shifted 
his interpersonal behavior in therapy. He was compliant (point 1) 



during the initial stage of treatment but became passively hostile and 
withdrawn in the second stage of therapy. Later changes in the treat- 
ment relationship can be similarly plotted — always in relationship to 
the average of the population. 

Previous Suggestions for a Two-Dimensional 
Classification of Personality Traits 

The notion of classifying human emotions in terms of four syste- 
matically related variables is certainly not novel. The history of 
psychology provides several interesting correspondences to the present 
system of arranging data in terms of the four nodal points. 

The four quadrants of the interpersonal system comprise blends of 
the nodal dichotomies: love versus hate and power versus weakness. 
The four "blended" quadrants fit rather closely the classical humors 
theory of Hippocrates. The upper left quadrant (hostile strength) 
equates with the choleric temperament, the lower left (hostile weak- 
ness) with the melancholic, the lower right (friendly weakness) with 
the phlegmatic, and the upper right (friendly strength) with the 

The same fourfold classification reappears in Freudian thought, 
Freud's treatment of the individual stresses two basic motives — love 
and hate. His theories of social phenomena and group interaction, on 
the other hand, emphasize domination, power, and the interaction of 
the weak versus the strong. In his open letter to Einstein "Why War?" 
these two avenues of Freud's thought intersect and illustrate his 
commitment to the four concepts. He presents his power theory first: 

Such then, was the original state of things: domination by whoever had the 
greater might— domination by hate violence or by violence supported by 
intellect. (2, p. 275) 

In the following paragraph he says: 

The situation is simple so long as the community consists only of a number 
of equally strong individuals. . . . But a state of rest of that kind is only theo- 
retically conceivable. In actuality, the position is complicated by the fact that 
from its very beginning the community comprises elements of unequal strength 
—men and women, parents and children— and soon, as a result of war and con- 
quest, it also comes to include victors and vanquished, who turn into masters 
and slaves. The justice of the community then becomes an expression of the 
unequal degrees of power obtaining within it; the laws are made by and for the 
ruling members and find little room for the rights of those in subjection. From 
that time forward there are two factors at work in the community which are 
sources of unrest over matters of law but tend at the same time to a further 
growth of law. First, attempts are made by certain of the rulers to set them- 
selves above the prohibitions which apply to everyone— they seek, that is, to go 


back from a dominion of law to a dominion of violence. Secondly, the op- 
pressed members of the group make constant efforts to obtain more power and 
to have any constant efforts to obtain more power and to have any changes that 
are brought about in that direction recognized in the laws— they press forward, 
that is, from unequal justice to equal justice for all. (2, pp. 276-77) 

Later, in the same paper, Freud goes on to summarize his familiar 
theories of individual motivation. 

According to our hypothesis human instincts are of only two kinds: those 
which seek to preserve and unite— which we call "erotic," exactly in the sense 
in which Plato used the word "Eros" in his Symposium, or "sexual" with a 
deliberate extension of the popular conception of "sexuality"— and those which 
seek to destroy and kill and which we class together as the aggressive or destruc- 
tive instinct. As you see, this is in fact no more than a theoretical clarification 
of the universally familiar opposition between Love and Hate which may per- 
haps have some fundamental relation to the polarity of attraction and repulsion 
that plays a part in your own field of knowledge. We must not be too hasty 
in introducing ethical judgments of good and evil. Neither of these instincts is 
any less essential than the other, the phenomena of life arise from the operation 
of both together, whether acting in concert or in opposition. It seems as 
though an instinct of the one sort can scarcely ever operate in isolation; it is 
always accompanied— or, as we say, alloyed— with an element from the other 
side, which modifies its aim or is, in some cases, what enables it to achieve that 
aim. Thus, for instance, the instinct of self-preservation is certainly of an 
erotic kind, but it must nevertheless have aggressiveness at its disposal if it is to 
fulfill its purpose. So, too, the instinct of love, when it is directed toward an 
object, stands in need of some contribution from the instinct of mastery if it is 
in any way to possess that object. The difficulty of isolating the two classes of 
instinct in their actual manifestations is indeed what has so long prevented us 
from recognizing them. 

If you will follow me a little further, you will see that human actions are 
subject to another complication of a different kind. It is very rarely that an 
action is the work of a smgle instinctual impulse (which must in itself be com- 
pounded of Eros and destructiveness). In order to make an action possible, 
there must be as a rule a combination of such compounded motives. This was 
perceived long ago by a specialist in your own subject, a Professor G. C. 
Lichtenberg who taught physics at Gottingen during our classical age-though 
perhaps he was even more remarkable as a psychologist than as a physicist. He 
invented a Compass of Motives, for he wrote. "The motives that lead us to do 
anything might be arranged like the thirty-two winds and might be given 
names on the same pattern: for instance, 'food-food-fame' or 'fame-fame- 
food'. So that when human beings are incited to war they may have a whole 
number of motives for assenting— some noble and some base, some of which 
they speak openly and others on which thev are silent. There is no need to 
enumerate them all. A lust for aggression and destruction is certainly among 
them: the countless cruelties in history and in our every day lives vouch for its 
existence and its strength. The gratification of these destructive impulses is of 
course facilitated bv their admixture with others of an erotic and idealistic 
kind." (2, pp. 280-82) 



The similarity between these suggestions for a "Compass of Mo- 
tives" and the circular classificatory system described in this book is 
so close as to require no further comment. 

In addition to these earlier approaches to a fourfold classification 
system of human motives, other similar conceptual schemes have been 
developed contemporaneously with (and independently of) the inter- 
personal system. 

The interpersonal system, it will be recalled, was developed from 
the rawest kind of empirical approach. It can be said that the patients 
in the earliest pilot study group developed the interpersonal circle by 
providing the varied pool of interpersonal responses which were 
gradually refined into the present circular continuum. It is most inter- 
esting, therefore, that the results of our empirical studies tend to con- 
firm hypotheticated fourfold classifications independently proposed 
by other writers. 

Ross Stagner, for example, in 1937 presented a two-dimensional 
representation of behavior which has a certain similarity to the inter- 
personal "compass." Stagner wrote: "The hypothesis which we wish 
to present is that the directions of variability in human behavior are 
very limited in number, present evidence suggesting that there are 
only two dimensions along which such variations may be plotted. 
These two dimensions may be considered: 1) approach to or with- 
drawal from a stimulus object; and 2) increased or decreased organis- 
mic activity with reference to the object." (II, p. 52) 

Although Stagner is noninterpersonal in his variable system and, 
perhaps, overly optimistic about the simplicity of direction and moti- 
vation, his paradigm attracts our interest for two reasons: First, it is 
remarkably similar to the interpersonal circular system. Secondly, it 
is close to the spatial theory of the genesis of interpersonal relations 
which we have discussed in the preceding pages. 

Talcott Parsons, who is perhaps the most sophisticated and syste- 
matically mature sociological writer of our generation, has described 
a conceptual method which he calls the "paradigm of motivational 
process." He states that this 

. . . started with the assumption that a process of interaction which has been 
stabilized about conformity with a normative pattern structure, will tend to 
continue in a stable state unless it is disturbed. Concretely, however, there will 
always be tendencies to deviance, and conversely these tendencies will tend to 
be counteracted by re-equilibrating processes, on the part of the same actor or 
of others. 

It was furthermore maintained that neither the tendencies toward deviance 
nor those toward re-equilibration, that is, toward "social control" could occur 
in random directions or forms. Deviance was shown to involve four basic 


directions, according to whether the need was to express alienation from the 
normative pattern— including the repudiation of attachment to alter as an object 
—or to maintain compulsive conformity with the normative pattern and attach- 
ment to alter, and according to whether the mode of action was actively or 
passively inclined. This yielded four directional types, those of aggressiveness 
and withdrawal on the alienative side, and of compulsive performance and 
compulsive acceptance on the side of compulsive conformity. It was further- 
more shown that this paradigm, independently derived, is essentially the 
same as that previously put forward by Merton for the analysis of social struc- 
ture and anomie. (9, p. 68) 

Thus, we see that two productive sociologists, Parsons and Merton, 
although working from somewhat different subject matters and frames 
of reference, have arrived at solutions for categorizing human inter- 
action which are close to the interpersonal circle. 

Another very interesting correspondence has developed from the 
researches of George T. Lodge. Lodge has developed some promising 
applications of the Haskell Coaction technique to psychological meas- 
urement. This is a method for plotting the resolution of two coacting 
variables in terms of a two-dimensional surface. The coaction compass 
functions exactly as the interpersonal circle, and the standard trigono- 
metric solutions of coaction variables have been applied by Haskell and 
Lodge to their data in the same manner as La Forge's formulas for 
the interpersonal system. 

Lodge describes his use of the Haskell Coaction Compass method 
as follows: 

The Coaction Compass as formulated by Edward F. Haskell is a general 
conceptual scheme which is beginning to find wide applications in biological 
and social science. This compass is a Cartesian coordinate frame strictly com- 
parable to the mariner's wind rose. Its use permits assignment of vector magni- 
tudes to the resultant forces from any two interdependent power systems, and 
their subsequent treatment by methods of analytic geometry. In the field of 
Clinical Psychology, it is convenient to view the processes of inhibition and 
facilitation as representing two such coacting power systems. It is not our 
purpose at present to go into the details of a coaction theory of personality as 
such. We have attempted a preliminary formulation of such a theory else- 
where. Here, we shall try only to set forth certain necessary steps for the 
interest of those who may wish to apply coaction reasoning in their analyses 
of Rorschach protocols. . . . 

The Rorschach method lends itself readily to the study of personality in 
terms of a coaction formulation, at least insofar as consideration of the scoring 
of determinants is concerned. If the form level of a response be regarded as 
reflecting the strength of the inhibitory process, and if the amount of expression 
of color, shading, and movement be regarded as reflecting the level of manifest 
affect or facilitation, the response may be represented geometrically as a re- 
sultant vector determined by the relative strengths of the two coacting power 
systems. (6, pp. 67h58) 


The Variability of Interpersonal Behavior 

Eitiploying the continuum of sixteen variables, summarized nu- 
merically, it is possible to make three different types of systematic 
studies of the same person. We can investigate the interpersonal be- 
havior of one individual at many levels of his personality. Charting 
the measurements for all aspects of behavior on the same circular grid 
provides a systematic pattern diagnosis of the structure of personality 
at one time. By adding summaries of the same measurements as they 
change in time, we obtain a picture of temporal variation in the 
multilevel pattern of personality. In the preceding example we have 
noted such a temporal change in one level of personality — inter- 
personal behavior in one cultural context, the psychoanalytic ses- 
sions. A third use of the circular continuum is to chart the varying 
patterns of behavior in different interpersonal situations. How does 
the patient behave with his boss, with his wife, with his children? 

These measurements of behavior, at different levels, at different 
times, and in different situations comprise the basic patterns and 
changing processes of personality. They are called structural, tem- 
poral, and situational variation patterns, respectively. Temporal varia- 
tion — the changes in personality patterns over time — has extreme 
functional importance since our prediction about future developments 
(e.g., prognosis for psychotherapy) is involved. Situational variation 
refers to the cultural relativity of interpersonal relationships. Struc- 
tural variation refers to the relationship among the levels of personal- 
ity and brings us to the basic issues of the notational system — the or- 
ganization of personahty into levels. 

The Formal Notational System: The Levels of Personality 

The fact that behavior exists at more than one level of awareness 
has been intuitively recognized for centuries. The discovery of un- 
conscious motivation — in the sense of a formal theoretical statement — 
was first made by Sigmund Freud. 

This was an epochal landmark in the study of personality and 
human nature. 

The neat personality structures of rationalistic psychology were 
exploded into an untidy disarray. It is no longer possible to depend 
on the solid validity of the subject's conscious report. If the subject 
in a perception experiment judges one stimulus object as larger than 
another, it may have to do with the physical aspects of perception — 
but it may also reflect a desire to agree or disagree with other subjects, 
to assist or frustrate the experimenter's purpose (as he imagines it to 


The concept of levels destroys the simple, unidimensional notions of 
behavior determined by chains of stimulus-response reactions. All the 
major learning theories since Freud, however cognitive and physical- 
istic they may strive to be, have by necessity taken into account this 
multidimensional quality of motivation. The complexity of human na- 
ture for the first time begins to command adequate conceptual respect. 

Accompanying the early positive rewards of the "unconsciousness 
theory" is a series of premature, intuitive concepts and logical falla- 
cies. To deal with some of these illogical procedures, we have stated 
in the fifth working principle that any statement about human be- 
havior must indicate the level of personality data to which it refers. 

When this postulate was applied to the varied mosaic of miscel- 
laneous protocols obtained from the pilot study cases, the first task 
required was to classify them into discrete levels. The questions then 
became: How many levels of personality should be employed? What 
are they? And how shall they be defined? 

Any solutions to these problems must be arbitrary, formal decisions. 
That is, we must assume no divinely instituted or platonically ideal 
number of personality divisions. In selecting the number of levels, 
we are limited on the broad side by the practicalities of the empirical 
method and on the narrow side by theoretical adequacy, that is (at 
this primitive state of our knowledge), if we have too many levels, the 
permutations and combinations of the interlevel relationships become 
impossibly unwieldy. If we have too few, important nuances become 
lost by being compressed into general categories. 

After reviewing the many types and sources of personality data, 
a classification into five levels was found to be the most effective. This 
decision is a notational procedure which seems to meet the functional 
criteria of the present time. When we say that it is convenient to con- 
ceive of five levels of personality, we do not imply that there is "really" 
or "eternally" such a structural division. Early psychoanalytic writers 
naively tended to imply, and the uncritical reader tended to assume, 
that there "really were" two or three levels of personality in the same 
sense that there "are" five fingers on the hand. When the formal na- 
ture of these divisions of consciousness was not made explicit, a meta- 
physical language threatened to develop. At this point we designate 
five levels of personality data which we suggest are the most profitable 
for research, theory, and functional prediction. 

These five general levels of personality data are: I. the Level of 
Public Communication; II. the Level of Conscious Description; III. 
the Level of Private Symbolization; IV. the Level of the Unexpressed 
Unconscious; and V, the Level of Values. These levels are defined in 


terms of the operations which produce the pertinent data. That is, 
the source of the data automatically determines the level of classifica- 
tion. In this way we obtain operational definitions of the five levels 
of personality. 

There are many different specific kinds of expression which can 
contribute data to any one level. For example, there are several ways 
in which fantasy symbols can be manifested — dreams, projective tests, 
fantasies, etc. All of these produce Level III data, although the opera- 
tions by which the themes are expressed are quite separate. In order 
to insure clarity and precision we always indicate (by code) the spe- 
cific source of the data. The general level is designated by a roman 
numeral and the sublevel operations are designated by a code letter. 
Level III-D, for example, means private interpersonal symbols ob- 
tained from dreams. Level III-T indicates private interpersonal sym- 
bols obtained from TAT stories. The general definition of levels and 
the specific test and rating procedures by which they are measured 
will now be presented. 

Level I (Public Communication) consists of the overt behavior of 
the individual as rated by others along the sixteen-point circular con- 
tinuum. These judgments are made by trained observers or by naive 
fellow subjects who observe the subject in interpersonal situations. 
They rate his interpersonal impact as it appears to them. What we ob- 
tain is a series of ratings of the interpersonal effect the subject has on 
others who share social situations with him. Other estimates of Level 
I behavior are obtained from special test procedures — situation test, 
prediction scales and the like. 

Level I data is objective or public — rather than private or subjec- 
tive. It may or may not agree with the subject's own view of the situa- 
tion. To obtain Level I data it is necessary to have the subject in- 
volved in social interaction and to have others rate their view of his 
purposive behavior. This gives a measurement of his social "stimulus 
value." Other specialized methods for assessing Level I require the 
patient to take criterion-specific tests (like the MMPI) which allow 
us to predict his interpersonal role. 

The situation in which we rate interpersonal behavior can be an 
extraclinic event or it can be restricted to the more controlled en- 
vironment of the clinic or assessment situation. The raters can be re- 
searchers, diagnostic or therapeutic clinicians, fellow patients, or 
family members. The meaning of the Level I rating thus depends on 
the cultural context and the category of the rater. These differences 
provide interesting sublevel variations of the broad, general Level I 
of Public Communication. 



There are five methods which provide estimates of Level I public 
behavior. These are coded as follows: 

Level I-Al: MMPI indices which reflect the interpersonal pressure 
exerted on the clinician by the patient's symptoms. 

Level I-R: Ratings by trained personnel of the patient's minute-by- 
minute behavior in a social situation. 

Level I-S: Sociometric ratings (from check lists) by fellow patients 
or by trained observers. 

Level I-P: MMPI indices wliich predict the interpersonal behavior 
to be expected in group psychotherapy. 

Level I-T: Scores from standard situational tests which assess the 
patient's interpersonal reactions. 

The following chapter is devoted to a detailed description of the im- 
phcations, measurement, and validation of Level I behavior. 

Level II {Conscious Descriptions) includes the verbal content of 
all the statements that the subject makes about the interpersonal be- 
havior of himself or "others." His descriptions of himself and others 
are obtained from a variety of sources — conversations, therapy proto- 
cols, autobiographies, check lists. They are then rated along the same 
sixteen-point circular continuum. We are interested here in the sub- 
ject's reported perceptions of himself and his interpersonal world. We 
are not interested at this level in the consensual accuracy of these per- 
ceptions or in the potential deeper meanings. We are concerned only 
with the phenomenological field — the way in which the subject re- 
ports his view of self and world. It must be noted that one single sen- 
tence expressed by a subject can provide both a Level I and a Level II 
rating. If a patient says, "I am a responsible person," the Level II rat- 
ing reflects the surface meaning of responsibility (coded as O) re- 
ported by the subject. Observers of the interpersonal context in which 
the sentence was uttered might agree that its Level I-R effect was to 
establish autonomy from the therapist (coded B) or superiority over 
other patients (also coded B). The reported self-perception usually is 
different from the interpersonal impact on or meaning to others. 

There are four methods which provide data for Level II descrip- 
tions of self and others. These are coded as follows: 

Level II-Di: Ratings by trained personnel of the verbal content from 
diagnostic interviews. 

Level II-Ti: Ratings by trained personnel of the verbal content from 
therapy interviews. 

Level II-C:, Scores from the Interpersonal Adjective Check List on 
which the patient checks his view of self and others. 

Level II-A: Ratings by trained personnel of the content of autobiog- 
raphies written by patients. 



The illustration, implications, use, and validation of this level of con- 
scious description will be considered in Chapter 8. 

Level III {Private Symbolization) consists of projective, indirect 
fantasy materials. These data come from a variety of sources — dreams, 
fantasies, artistic, or autistic productions, projective tests — which 
elicit imaginative expressions. The interpersonal themes of all these 
symbolic expressions are rated by two or more trained raters along 
the sixteen-point circular continuum. We thus possess a technique 
for systematically measuring the indirect autistic data of personality in 
terms of the same interpersonal variables which we use to categorize 
the public or conscious aspects of behavior. The broad general nature 
of the level categories must be mentioned again. There are many sub- 
level varieties of symbolic data. Some creative, projective tests, for 
example, may be more closely related to the level of conscious de- 
scription. Others may be consistently identified with the pattern of 
dream themes. The exact "depth" of any symbolic response depends 
on a variety of factors — cultural context, type of symbolic stimulus, 
the nature of the Level I behavior at the time, etc. The detailed sys- 
tematic organization and specific differentiation of these private pro- 
ductions becomes one of the most important and fascinating problems 
of current dynamic psychology. 

There are at present seven methods for collecting Level III pre- 
conscious data from patients. These are coded as follows: 

Level III-T: Ratings of TAT stories. 

Level III-IFT: Ratings from the Interpersonal Fantasy Test.^ 
Level Ill-i: Ratings of responses to the Iflund projective test. (3) 
Level III-B: Ratings of responses the Blacky projective test. (1) 
Level III-D: Ratings of interpersonal themes in dream protocols. 
Level III-F: Ratings of interpersonal themes from waking fan- 
tasies expressed by the subject. 
Level III-M: MMPI indices which predict to preconscious be- 

There is one distinction to be made in dealing with preconscious 
data that is most important. This is the division between the hero and 
the world personages in fantasy productions. Evidence from several 
samples suggests that clearly different sublevels of behavior are in- 

* The Interpersonal Fantasy Test is a Level III instrument developed by the Kaiser 
Foundation psychology research project to fit the interpersonal system. It is a TAT- 
type test in which the cards are designed to explore systematically the subject's fan- 
tasies about interpersonal relationships between heroes and paternal, maternal, cross- 
sex, and same-sex figures. Scores are obtained for Level III Self, Mother, Father, 
Cross-sex and Therapist. 


These findings are of considerable value because they define two 
distinct sublevels of the symbolic or preconscious area. One is desig- 
nated Level III Hero. This is the symbolic self-image. Its theoretical 
and clinical meaning is different from the preconscious images of the 
symboKc world. This latter area is designated Level III Other. These 
two subdivisions of symbolic expression have been found to be lawfully 
distinct. They often define different kinds of interlevel conflict and 
different personality types, and they are related to different sympto- 
matic pictures. Chapter 9 which is devoted to Level III symbolic be- 
havior will consider these distinctions. 

Level IV (the Unexpressed Unconscious) is defined by the inter- 
personal themes which are systematically and compulsively avoided 
by the subject at all the other levels of personality and which are 
conspicuous by their inflexible absence. Here we refer to those activi- 
ties which are consistently and deliberately "not present" in the per- 
sonality profile. These "unexpressed" aspects of personality are as yet 
unexplored. For this reason, this level will not be employed in the 
basic systematization that follows. 

The definition of Level IV is a problem as yet unsolved. The most 
convincing demonstration of the presence of motivation previously 
unexpressed (at the other three levels) would require two parallel sets 
of evidence. The negative proof would involve statistical demonstra- 
tion that the subject significantly avoids certain patterns of interper- 
sonal response with a frequency far beyond the expectations of 
chance. The proof positive requires that the same interpersonal themes 
be picked up in significant frequency by certain subliminal, indirect 
perceptual tests, e.g., abnormally long reaction times or perceptual 
distortions in response to thematic stimuli presented at spht-second 
(blurred) tachistoscope exposures. The implications and problems in- 
volved in the unexpressed behavior of Level IV will be surveyed in 
Chapter 10. 

Level V (Values) consists of the data which reflect the subject's 
system of moral, "superego judgments," his ego ideal. We refer here 
to the interpersonal traits and actions that the subject holds to be 
"good," proper, and "right" — his picture of how he should be and 
would like to be. These idealized interpersonal themes are obtained 
in the same manner as the conscious descriptions of Level II. We single 
out from interview, free association, check list, and questionnaire the 
expressions which concern his value-feelings. These are rated and 
scored according to the sixteen-point circular continuum. 

Like the other levels of personality, the "ego ideal" cannot be con- 
ceived of as a unitary or narrowly defined category. Some "values" 
may be consciously expressed — others may be rated as they appear 


in implied form. Thus some may be "deeper" than others. There are 
three methods for obtaining Level V ratings of the ego ideal. These are 
coded as follows: 

Level V-C: Scores from the Interpersonal Adjective Check List on 
which the patient checks his ego ideal. 

Level V-Di: Ratings by trained personnel of the subject's ideals as 
expressed in diagnostic interviews. 

Level V-Ti: Ratings by trained personnel of the subject's ideals as 
expressed in therapy interviews. 

The measurement and meaning of this level of behavior will be 
discussed in Chapter II. 

There follows in Table 1 a summary of the various sources of data 
for each level and sublevel of personality. We should observe again 
that the assignment of data to the appropriate level operates automati- 
cally. The source of the data routinely and rigidly defines the level. 
It should also be noted that while our method is rigid, behavior is 
flexible and fluid, and does not always follow our notational schemes. 
By this we mean that there exist sublevel variations; some Level II 


Operational Definition of Five Levels of 
Personality According to Source of Data 

Level I: (Public Communication) This level concerns the interpersonal impact of the 
subject on others— his social stimulus value. There are four different ways of ob- 
taining this measure: 

Level I-R: Ratings by trained personnel of the patient's minute-by-minute be- 
havior in a social situation. 

Level I-S: Sociometric ratings (from check lists) by fellow patients or by 
trained observers. 

Level I-M: MMPI indices which predict the interpersonal behavior to be 

Level I-T: Scores from standard situational tests which assess the patient's in- 
terpersonal reactions. 

I^evel II: (Conscious Descriptions) The subject's view of self and world obtained from 
interviews, autobiography, check list, questionnaire. There are four methods which 
provide data for this level: 

Level II-Di. Ratings by trained personnel of the verbal content from diagnostic 

Level II-Ti: Ratings by trained personnel of the verbal content from therapy 

Level II-C: Scores from the Interpersonal Adjective Check List on which the 
patient checks his view of self and others. 

Level II-A: Ratings by trained personnel of the content of autobiographies writ- 
ten by patients. 

Level III: (Preconscious Symbolization) The subject's autistic, projective fantasy pro- 
ductions. There are two sublevels of preconscious expression: Level III Hero and 
Level III Other. 


Level III Hero is defined by the interpersonal themes attributed to the heroes of 
preconscious protocols obtained from dreams, fantasies, projective stories 

Level III Other comprises the interpersonal themes attributed to the "other" 
figures from the same preconscious protocols. 

There are at present seven methods for collecting Level III data: 

Level III-T Ratings of TAT stories. 

Level III-IFT: Ratings from the Interpersonal Fantasy Test. 

Level III-i: Ratings of responses to the Iflund pro)ective test. 

Level III-B: Ratings of responses to the Blacky projective test. 

Level III-D- Ratings of interpersonal themes in dream protocols. 

Level III-F: Ratings of interpersonal themes from waking fantasies expressed 

by the subject. 

Level III-M: MMPI indices which predict preconscious behavior. 

Level IV: (Unexpressed Unconscious) This level is defined by two criteria: the inter- 
personal themes significantly omitted at the top three levels and significantlv avoided 
on tests of subliminal perceptions, selective forgetting, and the like. Specific methods 
for obtaining this data are not yet developed. 

Level V: (Ego Ideal) This level comprises the subject's statements about his inter- 
personal ideas, standards, conceptions of good and evil as obtained in interview, 
autobiography, questionnaire, or check list. There are three methods for obtaining 
Level V ratings of values: 

Level V-C: Scores from the Interpersonal Adjective Check List on which the 

patient checks his ego ideal. 
Level V-Di- Ratings by trained personnel of the subject's ideals as expressed in 

diagnostic interviews. 
Level V-Ti: Ratings by trained personnel of the subject's ideals as expressed in 

therapy interviews. 

reports (let us say from the intense confidence of psychotherapy) 
turn out to be much closer to our Level III measurements. Some 
symbolic productions (Level III) from subjects who are striving to 
"overload" their presentations in one thematic direction may duplicate 
Level II conscious reports. These sublevel shifts are generally due to 
differences in the social situation, or in the stimulus materials, or gen- 
eral variability factors such as time, oscillation, and interlevel dynamics. 
All of these are, fortunately, open to some systematic measurement 
and predictive control, and will be treated in a later publication. 

To conclude this preliminary glance at the five defined levels of 
personality, an illustration of the way data are assigned to levels may 
prove helpful. If a subject is rated as displaying aggressive behavior 
in a unit of interpersonal action, the rating of hostility (E) is then 
coded into the matrix of Level I-R variables. Should this same subject 
describe himself in a conscious report (on a check list) as friendly and 
agreeable, a Level II-C rating of affiliation (M) would be made. 
Should he report a dream in which the hero behaves in a submissive, 
trustful fashion, dependence (K) would be coded into the Level III-D 
pattern. Should nurturant behavior be absent from all of these three 
levels (to a statistically significant degree), and if it appears in the 


form of exaggerated avoidance or distortion of tenderness themes on 
Level IV measuring devices, then the presence of Level IV nurturance 
(N) could be inferred. If his description of his "ego ideal" on the 
check list stresses the themes of power and independence, then the 
Level V-C scores of A and B are emphasized. 

Let us assume that hundreds of additional measurements at all levels 
continue to emphasize the same pattern. The summary totals for each 
level are converted to standard scores, comparing them to the means 
of appropriate normative larger samples of cases. By means of the 
vector method described above we can chart the personality structure 
in the form of a diagram summarizing five levels of self-behavior. The 
data from each level has been converted into a systematic rating lan- 
guage which is standardized, and directly comparable with the data 
from other levels. The many implications and theoretical aspects of 
this multidimensional organization of personality will be discussed in 
Chapter 13, 

The Measurement of the Self-Other Interaction 

A final notationa^ procedure remains before the basic elements of 
the personality structure can be assembled. Formal recognition must 
be made of the fact that any interpersonal behavior involves more 
than one person — and by definition cannot be considered as an iso- 
lated phenomenon. We accepted (in the fourth working principle) 
the premise that the interpersonal theory logically requires that for 
each variable or variable system by which we measure the subject's 
behavior, we must include an equivalent set for measuring the parallel 
behavior of the subject's interpersonal world. 

The reciprocal nature of social interaction, the reflex way in which 
human beings tailor their responses to others, and the automatic way 
in which they force others to react to them will become one of the 
main points of emphasis in this book. To take systematic account of 
these interchanges (at all levels of personality) a notational step is 
required. This is accomplished by categorizing and summarizing 
separately the interpersonal responses of the subject and the specific 
others with whom he interacts. When we observe the subject's public 
communications at Level I we rate not only his purposive behavior, but 
also what others do to him. Then we score the patient's interpersonal 
responses to the psychotherapist and we also score the latter's reactions 
toward the patient. We note, for example, that the subject acts de- 
pendent (K) and the therapist reacts with nurturance (O). 

When we measure the subject's conscious reports at Level II, we 
rate not only his perceptions of himself, but also his descriptions of 
his interpersonal world as he views it. Thus we score the interpersonal 


themes the patient attributes to himself and, in addition, the themes 
he attributes to the specified "others" with whom he is concerned. We 
rate, for example, the subject's statement "I am helpless to solve this 
problem" (/) and his description of the therapist "You are a person 
who can help me with my problem" (O). When we summarize his 
Level II material, we obtain a numerical or diagrammatic total for 
the reported view of self, his view of his therapist, of his family mem- 
bers, of the other members of his therapy group, and all "others" he 
has described. 

When we deal with the symbolic data of Level III, we rate not only 
his fantasy themes attributed to self or to self-identified heroes, but 
also the interpersonal themes he assigns to the "others" with whom his 
fantasy self interacts. The subject might report, for example, a dream 
in which he attacks (E) his rejecting unsympathetic psychothera- 
pist (C). We summarize the Level III fantasy materials in the same 
manner — obtaining separate totals from his symbolic self and symbolic 

The usefulness of this self-other classification for the unexpressed 
themes of Level IV is, at present, an unsettled question. Some psy- 
chologists hold that the vague, diffuse themes from the less conscious 
areas of personality cannot be differentiated into self-other categories. 
Since there is no adequate data to settle this question, Level IV behavior 
will not be formally systematized in this book. 

The division of behavior into self-and-other does not seem to apply 
as directly to the "value" data from Level V. It might be assumed that 
the "ego ideal" or superego judgment of what's "right-and-good" 
holds as a general value system for one's view of self and all others. 
On the other hand, it is possible to obtain measurements on the "ideal- 
for-self" and the "ideal-for-specified-others." Thus the subject might 
be asked to describe his view of the "ideal" mother, the "ideal" father, 
the "ideal" spouse, the "ideal" therapist, the "ideal" boss, etc. The 
Kaiser Foundation research project is at this time conducting investi- 
gations of this sort, but the results are not yet tabulated. For this 
reason in this book. Level V will be considered as a unitary field and 
will not be divided into self and other. 

Variability Indices: The Organization of Personality 

The interpersonal system deals, therefore, with eight generic areas 
of personality data: two each (self and other) for Levels I, II, 
III, and one each for Levels IV and V. Since Level IV has been 
omitted from consideration in the current research, we shall be con- 
sidering in the following chapters seven generic areas of behavior. A 





Figure 4. Schematic Diagram Illustrating Seven Generic Areas of Personahtv at 
Four Levels and Listing Some Variability Indices of Personality Organization. 


preliminary diagram of personality structure can now be presented. 
Figure 4 illustrates the generic division of personality data with which 
we are concerned. Each of the seven circles denotes a generalized clas- 
sification of personality data. Inside each circle is printed a suggestive 
list of the sources of data for each level. This is a schematic, pictorial 
representation summarizing the notational procedures thus far out- 
lined. It is highly generalized. As it stands here it could not be used 
for clinical or research purposes since only one circle for "others" is 
represented. In practice we would have as many "others" circles as 
necessary to summarize the interpersonal behavior of each person with 
whom the subject interacts. 

More practical and accurate working diagrams are presented in the 
clinical and diagnostic chapters to follow. The stylized diagram in 
Figure 4 is presented to point up the next organizational issue, the re- 
lationship among the levels of personality. It will be noted that a series 
of lines connects the seven circles in Figure 4. These represent the 
dynamic interactions among the levels. Each circle, it will be remem- 
bered, summarizes the pattern of standard scores on the same matrix of 
sixteen variables. We can, therefore, make direct mathematical com- 
parison between levels. The line between Level II Self and Level III 
Self stands for the subtractive comparison between the two areas of 
personality, and indicates how similar or different they are. In addi- 
tion, it provides a numerical statement of those interpersonal themes 
which appear in private symbolization and which are not consciously 
attributed to self. These relationships among levels are called intra- 
personal variability indices. They are, in some respects, operational 
redefinitions of certain Freudian "defense mechanisms," since they 
systematically summarize the comparisons among the levels of per- 
sonality. The Freudians call these interlevel relationships "defense 
mechanisms" because they are seen as "warding off" instinctual im- 
pulses. We have, however, in principle hesitated to accept this focus- 
ing on the unconscious level of behavior and have accepted instead an 
emphasis on the over-all organization of all levels. We have tended to 
see these interlevel relationships simply as indices which reflect the 
structure of personality organization and the kind and amount of con- 
flict, or rigidity, or flexibility. 

We therefore call these relationships among the areas of behavior 
variability indices. They reflect in mathematical terms the tendency 
of any one level to dupUcate or balance the inevitable distortions of 
the other levels of personality. The definition, meaning, and function 
of these generic mechanisms of organization, will be discussed in Part 
III of this book. 



This chapter, by way of overview and prospectus, has presented a 
classificatory system for ordering interpersonal behavior. Five levels 
at which this behavior exists have been defined. A brief survey of the 
system of variability indices which link together these levels has been 

The subsequent chapters will be devoted to a detailed fiUing-in of 
the broad areas outlined in this chapter. The next six chapters will 
deal with the levels of personality — theory, measurement, method- 
ology, and meaning of the varieties of interpersonal behavior. Chap- 
ter 1 3 will deal with the variability dimension — and will present defi- 
nitions of the specific interlevel relations involved in the formal nota- 
tional system. 


1. Blum, G. S. The Blacky Pictures: A technique for the exploration of personality 
dynamics. New York. The Psychological CoqD., 1950. 

2. Freud, S. Why war? In Collected Papers. Translated by James Strachey. Vol. 
S. London: Hogarth Press and Institute of Psychoanalysis, 1950. 

3. Iflund, B. Selective recall of meaningless materials as related to psychoanalytic 
formulations in certain psychiatric syndromes. Unpublished doctor's dissertauon. 
University of Calif orma, Berkeley, 1953. 

4. LaForge, R., M. Freedman, T. Leary, H. Naboisek, and H. Coffey. The inter- 
personal dimension of personality: II An objective study of repression. /. Pers., 
1954, 23, No. 2, 129-53. 

5. LaForge, R., and R. Suczek. The interpersonal dimension of personahty. Ill An 
interpersonal checklist. /. Pers., 1955, 24, No. 1, 94-112. 

6. Lodge, G. T., and C. J. Steenbarger. Charting the course of the Rorschach inter- 
view. /. Gen. Psychol., 1953, 48, 67-73. 

7. Miller, Milton L. The traumatic effects of surgical operations in childhood on 
the interpretive functions of the ego. Psych. Quart., 1950, 20, 77-92. 

8. Naboisek, H. Interpersonal assessments of patients in group therapy. Unpub- 
lished doctor's dissertation. University of California, Berkeley, 1953. 

9. Parsons, T., and R. F. Bales. The dimensions of action-space. In Working 
papers in the theory of action. Glencoe, 111.- The Free Press, 1953. 

10. Reichenbach, H. R. The rise of scientific philosophy. University of California 
Press, 1951. 

11. Stagner, Ross. Psychology of personality. New York- McGraw, 1937. 


The Interpersonal Dimension of Personality: 
Variables, Levels, and Diagnostic Categories 


The following section of this book is devoted to a discussion of inter- 
personal behavior at five levels of personality. These levels are: 

I. The Level of Public Communication 

II. The Level of Conscious Communication 

III. The Level of Private Communication 

IV. The Level of the Unexpressed 
V. The Level of Values 

A chapter is devoted to each of these levels. Each chapter in- 
cludes a historical review of previous theories relating to the level in 
question, an operational definition of the level, a system for measuring 
behavior at the level, and a discussion of its significance. Considerable 
theoretical speculation as to the meaning of behavior at each level will 
be included. In order to preserve the descriptive and theoretical orien- 
tation of the book we have not included a detailed account of the re- 
search findings. Where there is evidence supporting these speculations, 
reference will be made to the scientific publication in which the perti- 
nent research has been described. 

This section is concluded by a description of the system of inter- 
personal diagnosis (Chapter 12). Here we employ the data from three 
levels of behavior to construct an objective multilevel diagnostic sys- 



The Level of Public Communication: 
The Interpersonal Reflex 

This chapter takes as its subject interpersonal communication. This 
aspect of personality, which we have designated Level I, is concerned 
with the social impact that one human being has on another. We shall 
consider first some methods for isolating and defining these interac- 
tions, and then proceed to their impHcations for personahty theory. 

The events studied at this level are the overt interpersonal activities 
of the individual. What a person does in any social situation is a func- 
tion of at least two factors, ( 1 ) his multilevel personality structure and 
(2) the activities and effect of the "other one," the person with whom 
he is interacting. 

In order to define and to discuss the level of overt communication 
it is necessary at times to tear it out of these two broader contexts in 
which it is always imbedded. The criteria of logical narration demand 
that we talk about interpersonal behavior in this chapter as though it 
exists apart from the other aspects of the person's personality struc- 
ture or apart from the behavior of others. These broader contexts are 
always implicitly referred to and should be kept in mind. 

Definitions and Illustrations 

The basic unit involved here is the interpersonal effect. We de- 
termine the interpersonal meaning of any behavior by asking, "What 
is this person doing to the other? What kind of a relationship is he 
attempting to establish through this panicular behavior?" The an- 
swers to these questions define the subject's interpersonal impact on 
the other one. For example, "He is boasting and attempting to estab- 
lish superiority"; or, "He is rejecting and refusing to help." 

We are concerned at this level with ivhat one person communi- 
cates to another. A father, for example, may employ one or one thou- 



sand words to refuse his child's request. The mode, style, and con- 
tent of the two rejecting expressions may be very different, but their 
interpersonal effect is the same — rejection. 

In studying the interpersonal purposes which underlie human be- 
havior, the following hypothesis has developed. It seems that in a 
large percentage of interactions the basic motives are expressed in a 
reflex manner. They are so automatic that they are often unwitting 
and often at variance with the subject's own perception of them. This 
facet of behavior is therefore a difficult one to isolate and measure. It 
is often unverbalized and so subtle and reflex as to escape articulate 
description. Sometimes these interpersonal communications can be 
implicit in the content of the discussion: Grandfather talks incessantly 
about the lack of energy and initiative of modern youth in order to 
impress others with the fact that he is a successful, self-made man. 
Grandmother talks incessantly about sickness, calamity, and death to 
remind others that the time may be short to repay her for the sacrifices 
she has made for her children. Grandfather never says openly, "I am 
better than you young people." Grandmother never says, "You 
should feel guilty and devoted to me." Grandfather's remark may be 
concerned with the issue of the 40-hour week. Grandmother may be 
quoting from the obituary column of the evening paper. Behind the 
superficial content of these expressions are the repetitive interpersonal 
motives — superiority and reproach. Behind the superficial content of 
most social exchanges it is possible to determine the naked motive com- 
munications: I am wise; I am strong; I am friendly; I am contemptuous; 
as well as the concomitant messages: you are less wise, less strong, 
likable, contemptible. Jung has described the "persona" as a mask-like 
front behind which more basic motives exist. The purposive behavior 
we are dealing with in this chapter is similar, but in emphasis something 
more important than just a social facade. It is closer, perhaps, to the 
"character armour" concept from the earlier writings of Wilhelm 
Reich, in that it assumes a major role in the personaUty organization. 
Its relationship to the "conversation of gestures" developed by Mead 
is, as we shall see, quite close. Let us examine some examples of Level 
I interpersonal communications. 

How A Poignant Woman Provokes a Helpful Attitude. A pa- 
tient comes to a psychiatrist for an evaluation interview. She reports 
a long list of symptoms — insomnia, worry, depression — and a list of 
unfortunate events — divorce, unsympathetic employer, etc. She cries. 
Whether her expressions are scored separately and summarized or 
judged on the over-all, we derive a clear picture of a JK approach — "I 
am weak, unhappy, unlucky, in need of your help." 


Let us shift now to the psychiatrist. He is under strong pressure to 
express sympathetic, nurturant communications. Helpless, trustful be- 
havior tends to pull assistance; that is, JK tends to provoke ON from 
the other one. Further, the patient-therapist situation is in essence one 
that lends itself easily to the "needs help-offers help" relationship. 
There exists a tendency for the psychiatrist to express openly (or 
much more likely, by implication) that he knows of a way by which 
the patient can be assisted. This may be communicated, not in ivhat he 
says, but in his bearing, attitude, his very quiet competence. 

What makes it more complex is the fact that the verbal expression 
may be quite different from the actual developing relationship. The 
psychiatrist may interpret the dangers of dependence and the necessity 
for self-help. The patient may agree. If both parties tend to over- 
emphasize verbal symbols, there may be an illusion that a collaborative 
relationship exists. Actually, the "nurturant interpreter-trustful fol- 
lower" situation still exists, not in what the participants are saying, 
but in what they are doing to each other. 

How THE Penitentiary Trains the Prisoner for Criminal Ag- 
gression. Many institutional or cultural situations have interpersonal 
implications so built into them that a flexible, collaborative relation- 
ship is impossible. In prison psychiatry, for example, as analyzed by 
Powelson and Bendix (8), it is virtually impossible for the doctor as 
well as for the patient to shake off the institution's implicit punitive 
contempt for the inmate. The penitentiary administration tells the 
prisoner, by the prison architecture, the structure of the guard-inmate 
relationship, and by every nonverbal cue possible that he is a danger- 
ous, evil, untrustworthy outcast. The prisoner often responds to this 
interpersonal pressure by accepting the role he is being trained for. 
That is BCD pulls EFG. The same interpersonal connotations were 
typical of the descriptive preanalytic psychiatry of the last century 
and of incarcerative psychiatry of the present. Here we must note 
again that human relationships are never one-sided and that those 
which are rigid or of long duration tend to be selective on both sides. 
Thus, as Powelson has pointed out, the recidivist criminal is least 
anxious when he is in passive rebellion against a strong punitive author- 
ity who feeds him and beats him. 

Interpersonal Implications Underlie Social Organizations. 
To a lesser degree any doctor-patient relationship tends to have pre- 
determined interpersonal structure. At least at the beginning it is 
highly loaded by the dependence-helpfulness axis. 

Unverbalized interpersonal assumptions tend to pervade every so- 
cial organization. The unwitting evaluation of the differing roles of 


orderly, nurse, psychotherapist, psychiatrist, and administrator in 
relationship to each other is an inevitable phenomenon in any psychia- 
tric hospital. The way in which eddies from these power whirlpools 
reach and relate to the patient is probably more important in terms of 
the remission rate than the number of electric shock machines or the 
skill of the psychotherapists. Factory, department store, office, uni- 
versity — all have these complex networks of routine, unverbalized 
evaluation through which power, prestige, contempt, punishment, ac- 
ceptance, etc., are expressed. Systematic understanding of these social 
hierarchies and their effect on the clients, patients, employees, cus- 
tomers, and students is a problem for the sociologist or the industrial 
psychologist. Investigations in these areas will very likely reveal that 
individuals tend to select jobs and occupational roles in accordance 
with their interpersonal techniques for anxiety reduction. We con- 
sider these phenomena here because they demonstrate the implicit and 
automatic nature of interpersonal reactivity. 

How THE Professor and Student Train Each Other To Be 
Professor and Student. The teacher-student relationship, obviously 
loaded with power implications, serves to illustrate some details of 
reflex communication. Professors are so addicted to the stereotyped 
teaching reflex that they often cannot inhibit the didactic response. 
We recall the psychology professor who had developed at some length 
in a lecture the thesis that teachers or psychotherapists should not give 
answers but should stimulate the student or the patient to seek answers 
himself. "Don't let them become dependent on you; make them think 
for themselves." As soon as the lecture was over, a graduate student 
(well trained to the dependency reflex) rushed up with a question: "In 
my undergraduate teaching section the students are continually ask- 
ing me to solve their personal problems and demanding answers. What 
shall I do?" Pausing only to clear his throat, the professor reflexly 
responded: "Yes, you'll always find your students tending to trap you 
into solving their problems for them — the problems that they should 
work out for themselves. Now what I'd do if I were you is, first, I'd 
get them to. . . ." The verbal content of an interaction can be quite 
divorced from the interpersonal meaning. 

These subtle, ubiquitous, automatic role relationships have as their 
function the minimization of anxiety. They set up smooth-flowing 
reciprocal interactions of ask-teach, attack-defend, etc. On those oc- 
casions when the pattern of interpersonal reflexes breaks down or is 
ambiguous, considerable distress generally results — manifested in the 
accustomed symptoms of anxiousness. Some students are made un- 
comfortable by a teacher who refuses to lecture and assume the 



authoritative role. Patients often manifest initial bewilderment and 
insecurity when the therapist appears disinterested in giving quick 
answers to their problems. Symbiotic marriage partnerships can be 
thrown into panic when the implicit assumptions of power, guilt, and 
dependence on which they rest are temporarily threatened. 

So far we have viewed interpersonal communications as automatic 
responses in standard institutional situations. We shall now proceed 
to study them in the context of the individual personality, in terms 
of the classification system of 16 variables presented in the last 

The preliminary data on which this system is based was obtained 
by giving extensive psychological test batteries to some 200 subjects 
and then recording their interactions in 45 discussion or therapy 
groups. The pretesting procedures produced many ratings of Level 
II (conscious perception) and Level III (symboUc productions) be- 
havior. The Level I-R communications were obtained by studying 
the subjects' behavior as they interacted with the four or five other 
group members. Let us select one subject as an example and follow 
him through this procedure. 

How A Sullen Patient Teaches Others to Reject Him. A 
thirty-year-old man came to the psychiatric clinic with complaints of 
depression, general immobilization, and social isolation. After intake 
interviews and testing, he entered a psychotherapy group along with 
four other patients. All the group members were strangers when they 



Figure 5. Summary of Interpersonal Interactions Between an Illustrative Subject 
and Four Fellow Group Members. 


The verbal transactions of the group were recorded and transcribed. 
Psychologists then rated each speech which this patient made and each 
verbal reaction by other patients to him. The sixteen-point circular 
continuum of variables (presented in the preceding chapter) was em- 
ployed in these ratings. The interpersonal actions of the subject and 
the reactions of the fellow group members during the first eight ses- 
sions were then summarized, combined into octants, and plotted on a 
circular profile. These Level I-R "self" and "other" profiles are pre- 
sented in Figure 5. 

This diagram tells us that the subject acted in a bitter, distrustful 
manner (FGH) in a group. He complained, demanded, accused, 
withdrew. His fellow patients reacted to him with a critical, un- 
sympathetic, rejecting exasperation (CDE). After eight sessions in 
the group the patient had virtually duplicated the suspicious, isolated 
pattern that had originally brought him to the cUnic. This rather pure 
and didactically simple interpersonal situation serves to illustrate sev- 
eral interesting aspects of interpersonal theory which will now be 

The Interpersonal Reflex 

First we ask, what did this patient do to get four strangers to agree 
on his social stimulus value? It seems that he trained them to react to 
him in a very specific way — provoking them to rejection and irrita- 
tion. This question becomes more important (from the diagnostic 
viewpoint) when we remember that he reports that over the span of 
his life he has consistently tended to remain isolated and despised by 
others. How does he do this? He made, on the average, ten verbal 
comments in each group session. What happened in these eight meet- 
ings of the group to bring about a significant disaffiliation? 

What Are Interpersonal Reflexes? When we trace his inter- 
personal actions back to the original recorded protocols we discover 
that a typical pattern of Level I interaction existed. The individual 
units of this behavior we call interpersonal ?nechanisms or interpersonal 
reflexes. They are defined as the observable, expressive units of face- 
to-face social behavior. 

These reflexes are automatic and usually involuntary responses to 
interpersonal situations. They are often independent of the content 
of the communication. They are the individual's spontaneous methods 
of reacting to others. 

The exact manner in which these Level I communications are ex- 
pressed is a complex problem. This much is clear: they are expressed 
partly in the content or verbal meaning of the communication, but 


primarily in the tone of voice, gesture, carriage, and external appear- 
ance. Although we do not know the specific method by which human 
beings communicate their emotional messages to each other, we can 
rate with reliable confidence the over-all, molar effect. Raters (trained 
psychologists or untrained fellow patients) can agree with impressive 
reliability in rating what subjects do to each other in interpersonal sit- 
uations. Preliminary research by Blanche Sweet (10) suggests that 
listening to recordings leads to more effective ratings than reading 
typed transcriptions. Sound movies would provide the optimal tech- 
niques for preserving the nuances involved in interpersonal reflexes. 
Future research may determine the specific way in which these spon- 
taneous interpersonal meanings manifest themselves to others. The 
reflex manner in which human beings react to others and train others 
to respond to them in selective ways is, I believe, the most important 
single aspect of personality. The systematic estimates of a patient's 
repertoire of interpersonal reflexes is a key factor in functional diag- 
nosis. Awareness and, if possible, modification of crippled or mal- 
adaptive reflexes should be a basic step in psychotherapy. When more 
evidence as to the mode of expression — gesture, carriage, content of 
speech — is at hand, some additions to therapeutic practice may 

The automatic and involuntary nature of interpersonal reflexes 
makes them difficult to observe and measure by a participant in any 
interaction. They are, for the same reason, most resistant to thera- 
peutic change. The more the members of the psychotherapy group 
tried to explain to the subject how and why he irritated them, the more 
he protested his feelings of injury. Later, intellectual insight and 
voluntary controlled changes to cooperative, self-confident behavior 
developed. These were, however, quite tentative and unnatural. Dur- 
ing many months of treatment spontaneous reactivity brought a 
return of the original responses. This involves, of course, the familiar 
process of "working through," basic to most therapeutic enterprises. 

Physiological and Interpersonal Reflexes. The on-going in- 
voluntary nature of these reflexes demands continual emphasis to keep 
them from slipping out of focus. This is the hidden dimension of be- 
havior. This is the area of personality which it never occurs to us to 
mention, so basic that it is taken for granted. Consider this analogy: 
A physician conducting a medical examination interview may ask the 
patient to report any physiological events he may have noticed during 
the previous day. The patient might describe the heavy feeling in his 
stomach after lunch, the headache during the evening. It would not 
occur to the patient to recall that he automatically blinked his eyes on 



the average of three times a minute, 180 times an hour, 2,880 times 
during the 16 hours of a waking day. Nor would the patient ordi- 
narily be able to report the absence of a reflex. He might describe the 
symptom that accompanies it, but the presence or absence of physical 
reflexes is generally unnoticed by the patient. His failure to mention 
them in the medical interview is, of course, natural and proper. It is 
not a sign of malignant "repressive" mechanisms but rather of the 
implicit nature of these important behaviors. 

Compare this, now, with the psychiatric interview. If asked to re- 
port the pertinent psychological events of the previous day, the patient 
might remember the feeling of depression in the forenoon, rage at the 
office, and worry over bills at home in the evening. It is inconceivable 
that he would or could recount that in almost every interpersonal situa- 
tion he conveyed by gesture, bearing, tone of voice, and the negativism 
of his verbalization a consistent message of pessimism and resentment, 
that over 70 per cent of his interpersonal mechanisms were in the same 
direction, that the "others" with whom he regularly interacts have 
been trained to respond to him in an irritated and rejecting manner. 
Nor would he indicate that the interpersonal reflexes expressing ten- 
der or affiliative purposes are crippled and inhibited. Again, his failure 
to mention these involuntary actions is not a symptom or a pathologi- 
cal repressive maneuver. The reflexes which we measure at Level I 
tend to operate as background to the verbal content of the communi- 
cation. It is the latter to which we consciously attend; but it is the 
former which set the tone and provide the interpersonal significance of 
the event. 

Interpersonal reflexes are considerably more tricky to deal with 
than their physical analogues. The medical examiner has routine, 
straightforward methods for checking physiological reflexes. The 
psychological situation is not so simple. The therapist may have to 
examine his own reflexive responses to the patient with great care 
before he can detect the exasperated boredom or irritation that this 
type of patient can pull from others. 

The Interpersonal Reflex Need Not Be Conscious. In this 
chapter we are dealing with the level of interpersonal action. In the 
following chapter we shall define conscious description as being a dif- 
ferent level of behavior. Level I is what the subject does. Level II 
is what he says he does. 

The interpersonal reflex is, therefore, not necessarily a conscious 
expression. It can be involuntary and not a deliberate or conscious 


This difference has been noted by other writers. Mead (7, p. 18) 
points out the difference between gestural behavior (Level 1) and 
consciousness (Level II). "The mechanism of the social act can be 
traced out without introducing into it the conception of consciousness 
as a separable element within that act; hence the social act, in its more 
elementary stages or forms, is possible without, or apart from, some 
form of consciousness." Cassirer (4, p. 53) has made the same distinc- 

Speech is not a simple and uniform phenomenon. It consists of different ele- 
ments which, both biologically and systematically, are not on the same level. 
We must try to find the order and interrelationships of the constituent ele- 
ments; we must, as it were, distinguish the various geological strata of speech. 
The first and most fundamental stratum is evidently the language of the emo- 
tions. A great portion of all human utterance still belongs to this stratum. But 
there is a form of speech that shows us quite a different type. Here the word 
is by no means a mere interjection; it is not an involuntary expression of feeling, 
but a part of a sentence which has a definite syntactical and logical structure. 
It is true that even in highly developed, in theoretical language the connection 
with the first element is not entirely broken off. Scarcely a sentence can be 
found— except perhaps the pure formal sentences of mathematics— without a 
certain affective or emotional tinge. 

The thesis of the present work is in agreement with these two 
authors. It is possible to express interpersonal behavior of which one 
is not aware. This is not to say that social reflex behavior is to be 
equated with the classic "unconscious." We are speaking instead of an 
involuntary, automatic behavior of which the subject can or cannot be 

Previous Literature on Interpersonal Communication. In 
making interpersonal communication a key concept in the present 
theory of personality, we are by no means introducing a new planet 
into the constellation of personality processes. The importance of 
reflex interactive behavior has been long recognized by sociologists 
and anthropologists. 

The psychologist-philosopher George H. Mead made a similar no- 
tion the kfeystone of his "social behaviorism." Mead has traced in great 
detail the development of human communication, and he discussed 
many concepts which are directly related to the interpersonal system 
developed in this book. While space does not permit the detailed 
analysis which Mead's work deserves, it might be useful to note some 
of the concepts which are directly related to Level I communications. 
Mead places the origin of communication in the "conversation of ges- 
tures" which, as he defines it, is very close to the definition of Level I 


used in this book. "We are reading the meaning of the conduct of 
other people when, perhaps, they are not aware of it. There is some- 
thing that reveals to us what the purpose is — just the glance of an eye, 
the attitude of the body which leads to the response. The communi- 
cation set up in this way between individuals may be very perfect. 
Conversation in gestures may be carried on which cannot be trans- 
lated into articulate speech." (7, p. 14) He continues in the same 
section to say: "But if we are going to broaden the concept of language 
in the sense I have spoken of, so that it takes in the underlying attitudes, 
we can see that the so-called intent, the idea we are talking about is 
one that is involved in the gesture or attitudes which we are using. 
The offering of a chair to a person who comes into the room is in itself 
a courteous act. We do not have to assume that a person says to him- 
self that this person wants a chair. The offering of a chair by a person 
of good manners is something which is almost instinctive. This is the 
very attitude of the individual. From the point of view of the observer 
it is a gesture. Such early stages of social acts precede the symbol 
proper, and deliberate communication." (7, p. 15) 

From this passage we see that the conversation of gestures (which 
we call reflex communication) is, for Mead, a lower order of behavior. 
As he develops his theory of the "significant symbol" he tends to de- 
preciate the importance of reflex, automatic (nonconscious) com- 
munication. When he compares it with vocal, self-conscious, reflexive 
language,^ this becomes quite apparent: "When, now, that gesture 
means this idea behind it and it arouses that idea in the other individual, 
then we have a significant symbol. In the case of the dog-fight 
(Mead's example of Level I, gestural communication), we have a ges- 
ture which calls out appropriate response; in the present case we have 
a symbol which answers to a meaning in the experience of the first 
individual and which also calls out that meaning in the second indi- 
vidual. Where the gesture reaches that situation it has become what we 

' Language usage becomes tricky at this point. When Mead uses the term reftexive, 
he means somethmg quite different from the term reflex as used in this book. He 
states: "It is by means of reflexiveness— the turning back of the experience of the 
individual upon himself that the whole social process is thus brought into the experi- 
ence of the individuals involved in it; it is by such means, which enable the individual 
to take the attitude of the other toward himself, that the individual is able consciously 
to adjust himself to that process, and to modify the resultant of that process in any 
given social act in terms of his adjustment to it. Reflexiveness, then, is the essential 
condition, within the social process, for the development of mind." (7, p. 134) This 
terminology contrasts with that used in this book. The interpersonal behavior at 
Level I which is (or at least, can be) nonconscious, involuntary, gestural, which 
involves an automatic communication with or "training of" the other one we call 
reflex. The variable by which we measure Level I behavior is the interpersonal reflex, 
or the interpersonal mechanism. 


call "language." It is now a significant symbol and it signifies a cer- 
tain meaning." (7, p. 45) 

Notice in this quotation how Mead distinguishes between the non- 
conscious language of gestures and the highly conscious significant 
symbol. The latter is a high-order concept and from the systematic 
point of view involves three separate levels of personality. Mead's 
purpose in developing a social theory of mind led him to employ com- 
plex combinations of personality variables. This is quite justifiable 
from the standpoint of Mead's conceptual intentions, but prevents a 
direct comparison to the systematic definition of levels which we are 
attempting in this book. In the broader sense. Mead's social behavior- 
ism can rightly be considered the creative watershed to which later 
theories of interpersonal relations can trace their sources. 

Roughly contemporaneous with Mead was another great pioneer in 
the field of culture and personality — Edward Sapir. Working as a 
linguist-anthropologist, Professor Sapir directly inspired many of the 
most well-known theories and investigations in the field of cultural 
anthropology. As early as the year 1927, Sapir was stressing the im- 
portance of interpersonal communication: 

If one is at all given to analysis, one is impressed with the extreme complexity 
of the various types of human behavior, and it may be assumed that the things 
we take for granted in our ordinary, everyday life are as strange and as un- 
explainable as anything we might find. Thus, one comes to think that the 
matter of speech is very far from being the self-evident or simple thing that we 
think it to be; that it is capable of a very great deal of refined analysis from the 
standpoint of human behavior; and that one might, in the process of making 
such analyses, accumulate certain ideas for the research of personality problems. 
There is one thing that strikes us as interesting about speech; on the one 
hand, we find it difficult to analyze; on the other hand, we are very much 
guided by it in our actual experience. That is, perhaps, something of a paradox, 
yet both the simple mind and the keenest of scientists know very well that we 
do not react to the suggestions of the environment in accordance with our 
specific knowledge alone. Some of us are more intuitive than others, it is true, 
but none is entirely lacking in the ability to gather and be guided by speech 
impressions in the intuitive exploration of personality. We are taught that 
when a man speaks he says something that he means to communicate. That, of 
course, is not necessarily so. He intends to say something, as a rule, yet what 
he actually communicates may be measurably different from what he started 
out to convey. We often form a judgment of what he is by what he does not 
say, and we may be very wise to refuse to limit the evidence for judgment to 
the overt content of speech. (9, pp. 892-93 ) 

Later in the same paper Sapir summarizes: "It should be fairly clear 
from our hasty review that if we make a level-to-level analysis of the 
speech of an individual and if we carefully see each of these levels in 


its social perspective, we obtain a valuable lever for psychiatric work. 
It is possible that the kind of analysis vi^hich has here been suggested, 
if carried far enough, may enable us to arrive at certain very pertinent 
conclusions regarding personality." (9, p. 905) 

If these predictions made over a quarter of a century ago seem most 
in tune with current trends, it can hardly be considered accidental. 
Working with Sapir at the time were many theorists who have since 
become well-known exponents of the culture theory of personality — 
Sullivan, Dollard, Thomas. 

The level of behavior which is operationally defined in this chapter 
as the Level of Public Communication possesses, therefore, a most 
eminent scientific heritage. Starting from Darwin and Wundt's con- 
cern with the gestural expression of emotion, taking its philosophic 
roots in the linguistic concepts of Sapir and Mead, and finding its 
psychiatric application in the writings of Fromm, Homey, Moreno, 
and Sullivan, the basic notion of interpersonal communication has for 
a century excited the interest of socially oriented theorists. 

The first clinical and empirical approach to interpersonal communi- 
cation was developed by J. L. Moreno. Many ingenious and creative 
innovations were introduced by this pioneering worker. For over 
twenty years Moreno has employed sociometric methods to study 
group structure. These techniques indicate the bonds of attraction 
and repulsion which exist among group members and provide an 
objective picture of the pattern of interpersonal relationships. 
Moreno's valuable contributions have not been fully exploited because 
of the absence of empirical studies. Moreno has not based his meas- 
urements upon a system of interpersonal variables. His sociometric 
methods possess considerable functional value, but they do not pro- 
vide an interpersonal diagnosis in terms of a fixed system of variables. 

In the last five years three comprehensive empirical systems for 
classifying interpersonal behavior have been described in the litera- 
ture. Bales ( 1 ) has presented a reliable and effective method of cate- 
gorizing interpersonal processes in terms of positive, negative, or neu- 
tral orientation toward a group goal. This has been applied mainly 
to group decisions and group problem-solving behavior. The English 
psychiatrist Bion (2) and his American follower Thelen (11) have 
developed a method of rating the individual's response to the group 
experience. This has been applied to problems of social structure in 
psychotherapy groups and to group-dynamics situations. The third 
method for measuring social interaction is the interpersonal system de- 
scribed in this book. The systems developed by Bales and Bion are 
major methodological achievements. They differ from the present 
interpersonal system in several respects. They are not tied to a theory 



of personality nor a system of multilevel measurement. The aim of 
Bales and Bion is to classify behavior that is most crucial to their par- 
ticular goals — group problem solving and group therapy process. The 
aim of the interpersonal system is to develop a method of measuring 
interpersonal behavior which will be coordinate with the measures of 
interpersonal behavior at other levels of personality and which will fit 
into a multilevel pattern of interpersonal diagnosis. The reflexes of 
Level I are, perhaps, the most crucial aspect of personality, but from 
the standpoint of functional diagnosis and dynamic theory design they 
must fit into a multilevel structure. 

In selecting the variables for classifying Level I communication, we 
have kept in mind, therefore, not the purpose or structure or task of 
the group, but the structure of the individual's total personality. With 
this discussion as background, we shall now consider the empirical 
methodology developed for classifying interpersonal behavior. 

Listing the Interpersonal Reflexes. To make objective meas- 
urements of the reflex phenomena of Level I, it was necessary to have 
a finite and defined list of interpersonal behaviors. Such a matrix has 
been presented in the form of the sixteen-point circular continuum. 
The problem becomes that of determining the various interactions 
which reflect the sixteen basic motivations. Because we are dealing at 
this level of personality with communication process — what one per- 
son does to another — it is convenient to use verbs (transitive verbs) 
as the descriptive terms. Figure 1, Chapter 6, presents the sixteen 
generic interpersonal themes along with a list of sample activities 
which illustrates the range of each point around the circle. 

For each generic theme there is, of course, an inexhaustible list of 
verbs. The terms used here are most appropriate for verbal exchanges 
in therapeutic or diagnostic contexts. Thus, we suggest that to boast, 
to act narcissistic ally, to establish autonomy and independence, to act 
self-confident all contain about the same proportion of dominance- 
hostility as indicated by the point B on the circle. By this we mean 
that they express the same qualitative purpose of narcissistic self-ap- 
proval. The fact that they difl^er in amount, degree, or extremity of 
the purpose is handled by the intensity scale. Other lists are necessary 
for categorizing nonverbal actions (frowns, gestures, voice tones) and 
preverbal situations (nursery school interactions, etc.). 

To illustrate this system of scoring social behavior, two samples of 
interaction are here presented: a section of a modern play, and a non- 
verbal nursery school exchange. A detailed description and illustration 
of the use of several Level I measures in the psychiatric clinic will be 
found in Appendix L 



The Scoring of Interpersonal Mechanisms as Applied to a 
Conversation in a Modern Play. The following passage represents 
a conversation among three of the central characters of Death of a 
Salesman by Arthur Miller.^ The scoring of each interpersonal reflex 
involves three items: the code letter representing the location of the 
action along the circular continuum, the verb considered most closely 
descriptive of the action, and the rating of intensity of the mechanism 
along the 4-point scale. The exchange between Biff and his mother, 
Linda, serves as a nice illustration of a punitive-guilty relationship. 

Linda: You're a pair of animals! Not one, not 
another living soul would have had the 
cruelty to w^alk out on that man in a 

Biff, not looking at her: Is that what he said? 

Linda: He didn't have to say anything. He was 
so humiliated he nearly limped when he 
came in. 

Happy: But, Mom, he had a great time with us. 

Biff, cutting him off violently: Shut up. 

[Without another word, Happy goes upstairs.] 

Linda: You! You didn't even go to see if he 
was all right! 

Biff, sail on the floor in front of Linda, the 
flowers in his hand; with self-loathing: No, 
Didn't. Didn't do a damned thing. How do 
you like that, heh? Left him babbhng in a 

Linda: You louse. You. 

Biff: Now you hit it on the nose! [He gets up, 
throws flowers in the wastebasket.] The 
scum of the eanh, and you're looking at 

Linda: Get out of here! 


tive verb 
























Condemn self 


Condemn 3 

Condemn self 3 


The Scoring of Interpersonal Mechanisms as Applied to Non- 
verbal Interaction in a Nursery School Situation. The next ex- 
ample of the scoring of interpersonal reflexes presents qualitative de- 
scriptions of the behavior of three children in a nonverbal nursery 

^ Arthur Miller, Death of a Salesman (New York: Viking Press, 1949), pp. 124- 
125; quoted by permission of the publisher. 



school situation. The reflexes are scored in the same fashion as the 
verbal interchanges described above. 

[Child A Is playing with a drum.] 

1. Child B Runs up and tries to pull drum 


2. Child A Tries to run away. 

3. Child B Trips A 

and pulls drum away. 

4. Child A Stays on ground, 

sobs loudly. 

5. Child B Parades with drum, pounding it in 

exhibitionistic manner. 

6. Child C Enters play area, walks to Child A. 

7. Child A Cries louder and pushes Child C 

away defensively. 

8. Child C Pushes Child A back and throws 

dirt on him. 

9. Child B Puts drum down and throws dirt 

on Child A. 


tive verb 



Takes by force 





Takes by force 











This interaction exemplifies the way in which individuals train 
others to reject and attack them. Notice how skillfully Child A (in- 
teractions 6, 7, and 8) pulls aggressive behavior from an initially well- 
intentioned sympathizer as well as from his original tormentor. 

As we observe from this last nonverbal interaction, the judgment of 
interpersonal reflexes is quite independent of the concrete medium of 
their expression. The mechanism of sympathize is scored for a non- 
verbal pat on the back as well as for a solely verbal reassurance — or 
from a combination of both. 

Five Methods for Measuring Level I Interpersonal Reflexes. 
In the preceding chapter it was pointed out that there are several meth- 
ods for obtaining Level I public communications. To insure clarity 
we have established the working rule that any mention of Level I be- 
havior must include a reference to the specific source of the data. 

When minute-by-minute ratings are made by psychologists of in- 
terpersonal behavior (either observed directly or derived from record- 
ings and transcriptions) the resulting data are assigned to Level 1-R. 

MMPI indices which reflect the interpersonal pressure generated 
by the patient's symptoms are coded Level I-M. 


MMPI indices which predict future interpersonal behavior in group 
psychotherapy are coded Level I-P. 

Scores from standardized situation tests which summarize the sub- 
ject's reactions are coded Level 1-T.^ 

When the subject's interpersonal role is summarized on the Inter- 
personal Check List by observers or fellow patients who have been in- 
teracting with him, the resulting sociometric indices are coded Level 

The Level I-R ratings of interpersonal reflexes are the basic meas- 
ures considered in the theoretical discussions of this chapter. In rou- 
tine clinical practice, however, we have found it necessary to rely on 
two sets of MMPI indices of Level I behavior. The Level I diagnosis 
discussed in the clinical chapters of this book is based on MMPI meas- 
ures of symptomatic behavior {Level I-M). The predictions of be- 
havior in group psychotherapy, derived from the MMPI, are labeled 
Level I-P. 

The reasons for employing these MMPI indices will now be dis- 

Level I-M Estimates of Symptomatic Behavior. Level I-R or 
Level I-S measurements are obtained from ratings of the subject's be- 
havior by others who have been interacting with him or observing his 
interactions. It is, by definition, necessary that the subject be involved 
in social relationships in order to make the Level I-R and Level I-S 
judgments. The ideal source of these ratings is the group therapy sit- 
uation where the subject's impact on several others can be determined. 

This poses a practical problem, however. The functional system of 
personality, which we are presenting in this book, is anchored to 
Level I-R and Level I-S. Since this is the "action level," we consider 
it to be the level of greatest immediate importance. In accomplishing 
interpersonal diagnosis'* at the time of intake into the psychiatric 
clinic. Level I-R and I-S ratings are generally not available. There has 
usually been no opportunity to observe the patient in extended inter- 
actions. In the Kaiser Foundation Clinic, the tests are generally ad- 
ministered after one intake interview with a clinician. For many 

^ The use of a situation test will not be illustrated in this chapter. The Kaiser 
Foundation project is now engaged in developing a standardized set of items for such 
a test. The test is being developed as a multiple-choice instrument and the patient's 
responses will be summarized and plotted in the same way as the other interpersonal 

* The system of interpersonal diagnosis is described in Chapter 12 of this book. 
Functional diagnosis is based on the multilevel interpersonal diagnosis and upon the 
diagnosis of variability (i.e., interlevel conflict). The use of the functional system of 
personality in accomplishing clinical diagnosis and prognosis is described in Chapters 
IS through 22. 



reasons it is impossible to get reliable Level I-R and I-S ratings from 
intake workers after one hour of interviewing. Thus, at the time the 
functional diagnostic system is called upon to make its predictions, it 
is forced to operate without its most important level of personality — 
the level of public communication. 

The ideal solution to this problem would be to develop methods for 
obtaining reliable estimates of Level I-S and I-R from the patient's 
interpersonal behavior at the time of intake interview and testing. 
Several factors — systematic and administrative — have made this pro- 
cedure infeasible. The problem of including estimates of the patient's 
purposive interpersonal impact at the time of intake diagnosis, as ex- 
pressed through his symptoms, has been met by developing psycho- 
metric indices. 

Every psychological symptom seems to have an interpersonal mean- 
ing, i.e., impHcations as to what the patient is communicating through 
the symptom, and what the patient expects to be done about it, etc. 
Symptoms are usually the overt reason for the patient coming to the 
clinic; they express an interpersonal message. 

In order to measure the symptomatic impact of the patient upon 
the clinician, we have combined eight MMPI scales into indices which 
can be plotted on the circular diagnostic grid. The Level I diagnoses 
employed in the research studies described in this book are based on 
these symptomatic indices. These measures are coded Level I-M. The 
MMPI formulas used to derive these indices are: vertical (dominance- 
submission) index = Ma -}- Hs — D — Ft; the horizontal (love-hate) 
index = Hy + K — F — Sc. These MMPI scale abbreviations and 
the methodology for measuring Level I-M are described in Appen- 
dix 1. 

These indices have considerable functional value for two reasons. 
First, they indicate the interpersonal reflexes employed by the pa- 
tient in approaching the clinic (as indicated by his motivation and the 
symptomatic pressure he exerts). In the clinical situation where a de- 
cision as to treatment and prognosis is the main functional issue, the 
interpersonal messages picked up by the MMPI predictive indices are 
exactly what the diagnostician must sense. The second value of these 
indices is that they are based on routine test procedures which are 
given at the time of intake evaluation. They provide necessary esti- 
mates of Level I-S behavior which would otherwise be lacking. They 
plug up a most crucial gap in the multilevel diagnostic pattern. 

The formulas for converting MMPI profiles into interpersonal 
measures are presented in Appendix 1 . The norms for converting the 
Level I-M dominance and love indices into standard scores are pre- 
sented in Appendix 5. 


Level I-P Indices for Predicting Interpersonal Behavior in 
Group Psychotherapy. In selecting patients for therapy groups and 
in planning the course of individual therapy, it is obviously helpful to 
have a foreknowledge of the expected interpersonal behavior. The 
Level I-M indices predict fairly well future behavior in group therapy. 
We have run several studies in which Level I-M ratings were corre- 
lated with sociometric ratings of Level I-S. These results are con- 
tained in another publication. (6) 

There were many cases, however, in which the Level I-M did not 
predict actual behavior in the group. The inaccurate forecasts were 
due to the fact that the group situation can pull responses from the 
patient that are different from his symptomatic behavior. A patient 
who is depressed and puts dependent pressure on the intake inter- 
viewer may become overconventional or helpful in his reactions to the 
future therapy group members. A different sublevel seems to be in- 
volved. The symptomatic indices seem to predict individual therapy 
(i.e., face-to-face reactions with a therapist) better than group ther- 
apy. This means that we can take an MMPI profile and calculate the 
Level I-M scores and forecast what the patient is going to do to the 
intake worker in planning treatment or to a future individual therapist. 
These MMPI cues work less well ifi forecasting what the patient will 
do to other patients in a group. 

For this reason a set of indices was devised which specifically pre- 
dicts behavior in group therapy. Two special MMPI scales for pre- 
dicting dominant or hostile roles were developed. These are labeled 
Level I-P. 

These scales were based on item analyses which studied the rela- 
tionship of Each MMPI item to Level I Sociometric indices of group 
patients. The MMPI indices which predict the patient's role in group 
therapy make it possible to plot on the diagnostic grid the patient's 
Level I-M score. This predictive index is useful in assigning patients to 
therapy groups. This is done as follows: We derive the Level I-P in- 
dices for all patients on the group-therapy waiting list and plot them 
on the same diagnostic grid. It is then possible to tell at a glance the 
range of role behaviors to be expected. The attempt is made to keep 
groups heterogeneous in respect to roles. That is, we do not want 
any group overloaded with one interpersonal type. Experience has 
demonstrated that a group comprised of hysterics will tend to manifest 
the same interpersonal reflexes, and interaction among patients is 
minimized. Similarly, a group with several passively resistant person- 
alities will tend to bog down into silence and mutual distrust. 

Selection of group patients is somewhat like the casting of parts in 
a play. We encourage lively interchanges among group members in 


which different role interactions develop. Predictive indices from 
MMPI scales which forecast expected behavior are the basis for as- 
signing patients to groups. They also assist the therapist by alerting 
him to the pressures which will develop. 

The use of MMPI predictive indices is illustrated in Appendix 1 and 
in Chapter 26. 

Routine Reflex Patterns. During any one day the average adult 
runs into a wide range of interpersonal stimuli. We are challenged, 
pleased, bossed, obeyed, helped, and ignored on an average of several 
times a day. Thus, the person whose entire range of interpersonal 
reflexes is functioning flexibly can be expected to demonstrate ap- 
propriately each of the sixteen interpersonal reflexes many times in any 

There are, however, many who do not react with consistent ap- 
propriateness or flexibility. One might respond to the pleasant as well 
as the rude stranger with a disapproving frown. Another might smile 
in a friendly fashion. If we study an extended' sample of a subject's 
interactions, an interesting fact develops. Each person shows a con- 
sistent preference for certain interpersonal reflexes. Other reflexes are 
very difficult to elicit or absent entirely. It is possible to predict in 
probability terms the preferred reflexes for most individuals in a spe- 
cific situation. A small percentage of individuals exist who get "others" 
to react to them in the widest range of possible behaviors and who 
can utilize a wide range of appropriate reactions. Most individuals 
tend to train "others" to react to them within a narrowed range of 
behaviors, and in turn show a restricted set of favored reflexes. Some 
persons show a very limited repertoire of two or three reflexes and 
reciprocally receive an increasingly narrow set of responses from 

Definition of Interpersonal Role. Most everyone manifests 
certain automatic role patterns which he automatically assumes in the 
presence of each significant "other" in his life. These roles are prob- 
ability tendencies to express certain interpersonal purposes with sig- 
nificantly higher frequency. The individual may be quite unaware of 
these spontaneous tendencies — to complain to his wife, to be stern 
with his children, to boss his secretary, to depend on the office man- 
ager. It must be remembered that we are talking here in statistical 
probability terms. The subject may have thousands of interactive ex- 
changes each day with each of his significant "others," and these may 
range all over the interpersonal continuum. When we obtain evi- 
dence that he consistently and routinely tends to favor certain mecha- 
nisms with one individual significantly more than chance and tends to 


pull certain responses from the other to a similar degree, then a role 
relationship exists. 

This selective process of employing a narrowed range of reflexes 
with certain "others" works, as we have seen, in a double reinforcing 
manner. Most durable relationships tend to be symbiotic. Masochistic 
women tend to marry sadistic men; and the latter tend to marry 
women who tend to provoke hostility. Dependent men tend to seek 
nurturant superiors, who in turn are most secure when they have 
docile subordinates to protect. 

The institutional role relationships, boss-secretary, prisoner-guard, 
student-teacher, etc., tend to be more stereotyped and fixed. Even 
so, there exists some room for role variability. Some secretaries 
"mother," nag, or even boss their nominal superiors. In general, how- 
ever, we can surmise that personality factors enter into the choice of 
occupation. Those people who are least anxious and most secure when 
they are submitting to and depending on strong authority tend to seek 
and hold subordinate jobs. The network of relationships even in the 
simplest office setup can be bewildering in its multilevel complexity. 
Even so, the institutional hierarchy patterns are less involved than the 
familial relationships. 

How A Pessimistic Man Reproaches His Wife. We cite here 
the very oversimplified example of the man who tends to complain to 
his wife. By this we mean that he reacts to his wife with the reflex 
of grumbling reproach (FG) with increased frequency, often to an 
inappropriate extreme. His voice may take on a tired, whiny quality 
the minute he enters the house. He can be, and often is, jolly, firm, or 
protective with his spouse. But as we pile up the thousands of inter- 
action ratings the trend towards mild complaint becomes increasingly 

Now we call these reactions reflexes because they are not deliberate 
or planned. He does not deliberately decide to inject the hurt, tired 
note in his voice. He does not plan the slight droop of the shoulders. 
He may not be aware of the continuous mild passive irritation. 

He may not even know the basic or broader reason for his bitter- 
ness. It might take some weeks of therapeutic exploration for him to 
verbalize his private feelings: (1) that he is a defeated genius whose 
failure was caused by his wife, (2) that he could be a success today if 
she had not persuaded him to marry and leave engineering school, ( 3 ) 
that he might be a rich man today if she had not persuaded him to give 
up that off^er in Texas, etc. More intensive analysis would, of course, 
allow the roots of these feelings to be traced back even further in the 
patient's history. 


Now this has been the oversimpliiied sketch of the interpersonal 
reflex repertoire of an essentially normal man and an essentially normal 
marriage. He is within normal limits because he maintains a reason- 
ably flexible range of interpersonal behavior. He probably can em- 
ploy the entire continuum of reflexes when appropriate, and without 
anxiety. If we could sum up all his interactions with all others over 
a period of time, we would see that all sixteen reflexes have been 
elicited, but that he tends to favor or overem.phasize the mechanisms of 
passive complaint and distrustful, realistic hesitancy. He can lead, he 
can express independence, he can support others, but he tends to a 
moderate but significant degree to favor a grumpy bitterness in his 
dealing with others. 

How He Provokes Superior Scorn from Four Strangers. This 
man entered a pilot study therapy group along with four other stran- 
gers. After eight sessions, a summary was made of his interpersonal 
behavior toward the other four group members. The same pattern 
was revealed. At times he lectured, argued, helped, cooperated, but 
the mechanisms which he spontaneously favored and manifested a sig- 
nificant majority of the time were those of passive resistance. At the 
same time, a summary was made of the interactions this man pulled 
from the others in the group, i.e., what they did to him. Again a fairly 
flexible pattern resulted. They listened to him with respect, they de- 
ferred to him, accepted his help, but the most frequent purposive be- 
havior directed toward him was a mildly critical superiority (BCD). 
They liked him, respected him, but on the whole felt moderate 
patronizing scorn in reaction to his grumbling approach. 

Notice that in seven sessions of brief interaction this subject (like 
the sample case described earlier) succeeded in duplicating his life 
situation with the four strangers in the group. This man, it must be 
remembered, is essentially normal. He can react flexibly and ap- 
propriately to most interpersonal situations. He has his favored inter- 
personal techniques for handling anxiety, as we all do, and this mild 
imbalance in the direction of grumpy pessimism gives him uniqueness 
and identity as a human being. His wife and his friends, very likely, 
understand and adapt to his gloomy realism with humorous (and 
sometimes irritable) impatience. 

A Masterful Defense of Sullen Distrust. A different picture 
develops when we turn to the maladjustive patterns. The suspicious, 
isolated, immobilized case described earlier in this chapter had an ex- 
tremely limited repertoire of social responses. He reacted in almost 
every situation with resentful distrust — to the group members, to his 
parents, to his acquaintances. It was next to impossible to elicit a sym- 



pathetic or nurturant response from this patient. These interpersonal 
reflexes were completely inhibited. The expressive behavior of this 
man was saturated with sullen, wary, growling distrust. By con- 
sistently exhibiting this narrow range of behavior in situations when 
they may or may not have been appropriate, he had trained the group 
members in the same manner that he had trained all the others in his 
life to condemn and isolate him. This patient's reflexive techniques for 
provoking rejection were so well developed that the most well-inten- 
tioned, friendly approach made no dent in his armor. However sym- 
pathetic the "other one" might be, his masterful, consistent sullenness 
would stimulate eventual impatience. This inevitable, exasperated re- 
jection, of course, increased his aggravation and would tend to in- 
crease the probability of further isolation. This reciprocal process by 
which human beings tend to pull from others responses that tend to 
maintain their limited security operations will be discussed under the 
headings of 'The Principle of Self-Determination" and "The Prin- 
ciple of Reciprocal Interpersonal Relations." Before treating these 
issues, we shall pause to consider some of the methodological impli- 
cations of the reflex behavior of Level I. 

It will be recalled that the Level I measure is always in terms of a 
rating of the subject's interpersonal behavior by someone else. The 
subject must be involved in a social situation. The Level I judgments 
are then made by the observers — psychologists or fellow participants 
in the interaction. 

Patients as Diagnostic Instruments. Our measuring instrument 
is, therefore, another human being. Since interpersonal behavior is a 
functionally important dimension of personality, it is quite natural that 
we measure it directly — in terms of the actual social impact that the 
subject has on others. Some interesting implications develop. By al- 
lowing the patient to react with others — say in a group therapy situa- 
tion — we make it possible for him to demonstrate, directly and openly, 
his repertoire of interpersonal reflexes. He tends to recreate to a mild 
extent in the group his neurotic adjustment. He accomplishes his 
own interpersonal diagnosis. 

The therapeutic group, thus, serves as a small subsociety, a minia- 
ture world. The members of a therapy group have a valuable diagnos- 
tic function. When we ask them to rate each other's interpersonal be- 
havior (on a check list or sociometric blank, covering the range of the 
sixteen generic variables) we obtain an estimate of what each patient 
has done to the others. 

Why Patients Produce Better Interpersonal Diagnoses than 
Psychologists. Patients tend to rate each other much in the same 



way as trained psychologists. Sometimes the members of a therapy 
group see a fellow patient differently from the therapist. When these 
discrepancies in Level I ratings occur, it is usually the psychiatrist or 
psychologist whose judgments are less accurate and less valuable. 
Clinicians' judgments of patients tend to be complicated affairs. They 
are often very derived, distorted by theoretical or "depth" considera- 
tions. Naive, untrained subjects — fellow patients, family members — 
generally judge each other in terms of their direct reactions to the sub- 
ject. They tend to like, fear, respect each other, and their ratings 
reflect these reactions. They do not "psychologize." The ratings of 
trained professional workers tend to be much more intellectual. It is 
not good form for them to admit that they like, fear, or look up to a 
patient. Their ratings, indeed, are classically supposed to be divorced 
from these personal reactions. 

We have found, therefore, that psychologists and psychiatrists tend 
to give interesting and theoretically valid ratings of deeper motives and 
future developments. Since patients spend most of their time inter- 
acting with nontrained, psychologically naive individuals, it is from 
the latter that we obtain the best diagnosis of their "main street" 
stimulus value. The meaning of such a rating is not clear-cut. Many 
extraneous factors influence it — the personality of the rater and the 
climate of the therapy group, to name two. Some of these complica- 
tions can be handled by standardizing procedures and corrections for 
perceptual distortion on the part of the rater. Others cannot be con- 
trolled. Even so, the rating from the nonprofessional is much more 
straightforward an estimate of Level I communication. 

Professional Clinicians as Measuring Instruments. There are 
very good reasons why trained clinicians are less effective and reliable 
when asked to rate single-level variables. The factors that make them 
good clinicians tend to hamper their becoming good rating instru- 
ments. The key to clinical skill is the ability to make multilevel ob- 
servations and to synthesize them. The good psychiatrist is trained to 
perceive many cues from many levels at one time and to act upon 
these cues effectively. 

He is able to predict what the multilevel behavior of the patient 
will be in the long-range future, and also in the immediate future. 
Thus he develops the intuitive expectation as to the patient's reac- 
tion to an interpretative intervention on his part. He is able to grasp 
what the patient may be communicating at several levels as he free- 
associates. He attends, now to the verbal content, now to the sym- 
bolic cues (slips of the tongue, etc.), now to the immediate inter- 
personal pressure from the patient. His behavior in response to the 


patient's activity is usually based on a complex integration of these 
many cues. In many cases he does not stop to sort out all these cues 
into categories. He may be hard put to explain exactly what it was 
in the pattern of the patient's communication that led him to his con- 

Now this is a most frustrating situation to the scientist who seeks 
to measure factors involved in therapeutic communication or diag- 
nosis. Scientific ratings are supposed to be reliable, repeatable, ex- 
plicit, specific. At this primitive stage of the science they are in- 
evitably single-level measurements. Often the scientist is very dis- 
couraged when he attempts to pin the clinician down. He knows that 
the clinician responds to cues with considerable skill. But his rating 
scales fail miserably to tap the richness of the clinical experience. The 
practitioner is likewise frustrated and sometimes irritated by what 
he calls the "simple-minded" quantitative approach of the scientist. 

The Kaiser Foundation project has worked out a tentative solu- 
tion to this dilemma by avoiding the use of clinicians as rating instru- 
ments. The clinician's insights are employed in setting up the system, 
in determining the rating categories. Thus we have long discussions 
with psychiatrists and practitioners before deciding how many levels 
to employ in the system. Clinicians tell us what kind of conflicts, iden- 
tification patterns, and therapeutic phenomena they run into. They 
produce many multilevel hypotheses about personality dynamics, the 
nature of change in therapy, and so on. The system is then expanded 
and revised to get at these phenomena. New experimental uses for the 
system are suggested. A clinician may report that certain patients 
show a particular imagery of "Father." The system is then broadened 
to get at this dimension — we may add a new rating category to the 
TAT analysis which picks up "symbolic view of Father." Clinical 
intuition is thus mainly responsible for what kind of a system and what 
kind of measurements are made. 

But we attempt to keep the clinician far away from the actual 
measurement process. He tells us what to measure; but we do not ask 
him to measure it for us. At the present time the project's measure- 
ment procedures (both research and diagnostic) are executed com- 
pletely by nonprofessional workers. The tests are administered by 
trained technicians. The specific ratings are done by the patient about 
himself, by his fellow patient, or by technicians carefully trained to 
make unilevel judgments. The TAT stories are not employed as global 
productions upon which multilevel analyses of the patient's personality 
can be based. They are defined as Level III data. A crew of intelli- 
gent, but nonpsychologically trained technicians then moves in to rate 
the TAT stories for the interpersonal themes. These technicians are 


not encumbered with the complex cHnical skills or broad theoretical 
conceptions. Their job is to do unilevel ratings, which they accom- 
plish with straightforward competence. 

The standard interpersonal system test battery includes seven dif- 
ferent tests. These are administered, scored, rated, and profiled by 
nonprofessional or semiprofessional help. In essence, the hundreds of 
molecular scores are fed into the system and the resulting matrix of 
multidimensional scores is handed to the clinician who then interprets 
it. The clinician applies his creative, intuitive skills to understand the 
complex patterns of scores and to relate this to the facts that he has 
about the patient from the interview. He makes sense out of a pattern 
of scores — a task which neither the unsophisticated patient nor the 
psychologically untrained technician can hope to perform. The pro- 
fessional energy is thus applied to developing the machinery of the sys- 
tem and to the final product which comes from this machinery. The 
running of the machine and the processing of the measurements (in- 
cluding ratings of symbolism) are accomplished by specially trained 

There is one occasion upon which we ask clinicians to make ratings. 
This is done when we want to study the clinicians and not the patients. 
If we ask twenty clinical workers to rate a group of patients or a set 
of test scores on the variable repression or ego strength, the results tell 
us how the individual clinician or how the entire group of clinicians 
conceive of these two variables. In two research explorations done by 
the research group it was determined that the clinical psychologists 
who rated repression and ego strength relied mainly on Level II cues. 
Patients who claimed to be strong, friendly, and healthy were rated 
as repressers and having strong egos. The results of the ratings thus 
told us how these psychologists conceived of the variables and did not 
necessarily measure the variables in an independently vaUd manner. 

When we ask untrained people for unilevel ratings, we have a fair 
idea of the meaning of the data. When we ask clinicians to make uni- 
level ratings, we are misusing their complex skills, confusing the mean- 
ing of the system, and in most cases lowering reliability. 

The Principle of Self -Determination 

In the preceding pages as the illustrative case material has unfolded, 
I have consistently employed a rather cumbersome circumlocution to 
describe the interaction between the sample subject and the "others" 
with whom they interact. Most statements describing what "others" 
did to the sample case were worded so as to give responsibihty to the 
subject. Thus we say, "He trained or provoked the group members 
to reject him," rather than "They rejected him." In the listing of il- 


lustrative interpersonal reflexes (Figure i), it may have been noted 
that both active and passive phrases were used. Thus for the inter- 
personal reflex G we have included acts rejected and provokes rejec- 
tion. We take the subject as the focus of attention and as the locus of 

I have tried to stress the surprising ease and facility with which 
human beings can get others to respond in a uniform and repetitive 
way. Interpersonal reflexes operate with involuntary routine and 
amazing power and speed. Many subjects with maladaptive inter- 
personal patterns can provoke the expected response from a complete 
stranger in a matter of minutes. The defiant chip-on-the-shoulder; 
docile, fawning passivity; timid, anxious withdrawal — these are some 
of the interpersonal techniques which can pull the reciprocal reaction 
from the "other one" with unfailing regularity. Severe neurotics — de- 
fined at this level as individuals with limited ranges of reflexes — are 
incredibly and creatively skilled in drawing rejection, nurturance, etc., 
from the people with whom they deal. In many cases the "sicker" the 
patient, the more likely he is to have abandoned all interpersonal tech- 
niques except one — which he can handle with magnificent finesse. 
Most clinicians who have dealt with the disorder will be glad to testify 
that the so-called catatonic negation is a powerful interpersonal ma- 

Assigning the causative factor in interpersonal relations to the sub- 
ject is a standard procedure in dynamic psychiatry. The skillful 
therapist is usually not inclined to join the abused, unhappy, masochis- 
tic patient in lamentation. He is much more inclined to ask himself 
and eventually the patient, "What do you do to people with con- 
sistent and consummate skill to get them to beat you up?" The prin- 
ciple involved here holds that interpersonal events just do not happen 
to human beings by accident or external design. The active and execu- 
tive role is given to the subject. 

This principle (as is the case with most other psychological con- 
cepts) has been described and given more eloquent expression by 
novelists. Here, for example, is D. H. Lawrence outlining the notion 
of self-determinism: "No man . . . cuts another man's throat unless 
he wants to cut it, and unless the other man wants it cut. This is a 
complete truth. It takes two people to make a murder: a murderer 
and a murderee. And a murderee is a man who is murderable. And 
who is murderable is a man who in a profound if hidden lust desires 
to be murdered." (5, p. 36) 

Human Beings Resist Taking Responsibility for Their Situa- 
tions. This point of view plows headlong into the most widespread 


resistance. It threatens the most cherished beliefs of Western philos- 
ophy — from Sophocles (who stresses fate) to the modern mental 
hygienists (who overemphasize parental behavior) . What is more im- 
portant, it threatens the most cherished illusions of the average man 
who bases his security and self-esteem on the traditional procedure of 
externalizing blame. 

What we are saying here to the human being is, "You are mainly 
responsible for your life situation. You have created your own world. 
Your own interpersonal behavior has, more than any other factor, de- 
termined the reception you get from others. Your slowly developing 
pattern of reflexes has trained others and yourself to accept you as 
this sort of person — to be treated in this sort of way. You are the 
manager of your own destiny." 

This attribution of responsibility to the subject we have called the 
Principle of Self-Determination. Although it has the deceptive ap- 
pearance of simplicity, it is, on the contrary, the most complex kind of 
concept. To this notion of self-determination the average person is 
willing to give halfhearted and halfway approval. It is easy to see 
where the successful person can be self-made. He chooses his goals, 
works for them, and makes the grade. It is accepted quite naturally 
that men strive and bargain for the interpersonal goals reflected in one 
half of the spectrum — independence, power, popularity, affection (i.e., 
B, A, P, O, N, M). It is often less comprehensible that men should 
actively seek the interpersonal states represented by the other half of 
the circular continuum — dependence, weakness, distrust, and self- 
effacing modesty (E, F, G, H, I, J). People, it is held, just don't seek 
to defeat themselves. 

How Three Human Beings Got What They Bargained for. 
A patient poignantly reports: "What I want more than anything else 
is to marry a dependent, feminine girl, but my three ex-wives were 
bossy, exploitive tyrants." This man may at the level of conscious 
awareness "want" a feminine girl, but his Level I behavior — im- 
mobilized, distrustful, and masochistic — is enough to force the most 
neutral woman into exasperated activity. 

Another patient states: "What I want in a husband is a strong, suc- 
cessful man who will take care of me; but all I seem to attract are 
penniless artists and passive, dreamy bookworms." This woman may, 
at Level II, consciously wish for a strong husband; but her mothering, 
responsible Level I reflexes are so automatic and deeply ingrained that 
the strongest man would feel smothered and alienated by the ma- 
ternal stability, to which dependent men are drawn with moth-like 


A third patient says: "I want more than anything else to finish my 
college training and get established in a profession, but all the profes- 
sors I have studied with are narrow-minded men who reject my ideas 
and end up by flunking me." This man may wittingly desire the 
prestige of professorial responsibility, but his rebellious, defiant re- 
flexes eventually exasperate and frustrate even the most sympathetic 
mentor. What human beings consciously wish is often quite at vari- 
ance with the results that their reflex patterns automatically create for 
them. For these people the sad paradox remains that voluntary inten- 
tions, verbal resolutions, and even intellectual insight are operationally 
feeble and numerically infinitesimal compared to the ongoing 24-hour- 
a-day activity of the involuntary interpersonal reactions. The frus- 
trated student just mentioned may in a burst of intellectual awareness 
decide to conform to the academic demands and return to college. 
This resolution is a conscious, voluntary efl"ort — very much like prac- 
ticing for an hour to prevent the eye-wink reflex from operating 
when an object is waved in front of it. Such conscious control cannot 
be maintained 24 hours a day or he would be able to concentrate on 
nothing but the eyelid reflex. Analogously, our rebellious student 
faces the difficult task of fighting a continuous, exhausting battle 
against his spontaneous tendencies (1) to sneer and balk at authori- 
ties, thus (2) pulling from them an eventual impatient rejection, which 
(3) increases his tendency to sneer and balk. 

Why Human Beings Develop the Reflexes of Weakness and 
Rejection. A second logical objection to the concept of interpersonal 
reflex looms up here. The eye-wink reflex, it might be argued, is 
naturally acquired, universal to all men, and survivally favorable. The 
patterns of social reactivity cited here meet none of these criteria. 
What is the rationale which explains how different human beings de- 
velop different rigid, self-defeating techniques of adjustment? 

The first point to note is that we are concentrating here — for the 
sake of illustration — on maladaptive phenomena. The description of 
these extreme reflexes gone wild implies that other reflexes are, in 
contrast, inhibited. The masochistic man could not maintain a mini- 
mum of independent assertiveness; the maternal role prevented the 
responsible woman from manifesting the reflexes of docility and trust; 
the rebel possessed conformity and affiliation reflexes which were qui- 
escent or extinct. 

In the adjusted, well-functioning individual, the entire repertoire 
of interpersonal reflexes is operating spontaneously, flexibly, and ap- 
propriately — and when the survival situation demands aggression, he 
can aggress; when it calls for tenderness, he can be tender. Human 



societies, however, tend not to be too well balanced. They tend to 
put a premium on certain interpersonal responses — competitiveness or 
slavish submission, for example. To survive and flourish, human beings 
must tailor their responses to the demands of such imbalanced cultures. 
Even in the most heterogenous and tolerant society the developing 
personality interacts with so many inflexible pressures (e.g., parent's 
personalities, subcultural demands) that a hierarchy of preferred re- 
flexes develops. To say that human personality is varied and different 
is to say — at this level — that most everyone tends to overemphasize 
certain automatic interpersonal responses and to underemphasize 

The questions still remain: Why do human beings limit their ma- 
chinery of social adjustment, manifest narrowed spectra of reaction, 
and provoke a restricted set of reactions from others? Why do some 
individuals have no ability for realistic, modest self-criticism (H) and 
compulsively express only narcissistic self-enhancing mechanisms (B) ? 
Why do others cling to retiring modesty and eschew the responses of 
proud self-confidence? Most puzzling of all (to the occidental mind) : 
Why do some of our neighbors masochistically court interpersonal 
humiliation — doggedly provoking rejection and isolation from others? 

For the answers to these questions we return again to Sullivan. He 
defines personality as the pattern of interpersonal responses employed 
to reduce anxiety, ward off disapproval, and maintain self-esteem. As 
the individual develops, he discovers that certain interpersonal re- 
sponses bring danger; some bring a narrow, uncomfortable, but cer- 
tain security. To use others would involve broad, attractive, but 
conflictful uncertainty. The more anxiety-provoking the individual's 
world — particularly his parental home — the more likely he is to select 
the familiar, narrow, certain, lesser anxiety and to avoid the promising 
but uncertain potentialities. The basic meanings behind any personal- 
ity pattern are difficult to evaluate. The complex behavior of counter- 
poised motives at difl'erent levels creates the appearance of a terribly 
anarchic system. It is clear, however, that for many people self- 
esteem and security involve surprising maneuvers — including extreme 
self-punishment (at Level I). The "search for suffering" (H) can 
have an inexhaustible number of meanings — all functions of the multi- 
level integration of personality. Getting her husband to beat and 
exploit her can allow the masochist externalization of guilt, propitia- 
tion of guilt, passive expression of hostility, and the intense pleasures 
of narcissistic martyred self-pity. "No one suffers more than me; 
watch and I can prove it." The projection of blame for failure on 
others, of course, requires more and more demonstration of failure 
in order to maintain the allegation. In addition, by selecting this set of 


aggression-provoking responses, the masochist avoids the potentially 
conflict-laden area of active hostility (DE), marital collaboration and 
-sexual partnership (LM), and assertive responsibiUty (BAP). Most 
severe martyrs, of course, express indirectly the purposive behaviors 
that they inhibit at the public level. They accomplish their private 
aims — aggressive, narcissistic, exploitive, and, very likely, sexual — by 
the indirect, cumbersome, and unsatisfactory method of acting hurt 
and provoking aggression from others. 

The Pressure to Repeat Responses. Interpersonal activities are 
designed to avoid the greater anxiety. It might be said in general that 
the human being experiences less anxiety in a familiar situation than 
in a strange one, and less anxiety when he is employing familiar 
responses than strange ones. Reciprocal relationships with crucial 
"others" develop quite naturally here. The more an individual re- 
stricts his actions to one narrow sector of the interpersonal spectrum, 
the more he restricts the social environment he faces. That is, the man 
who continually employs submissive reflexes tends to train people to 
boss him and discourage people from looking to him for forceful 
leadership. This tendency to repeat the patterns of the past is similar 
to the principal of least action which is described by Whitehead (12, 
p. 108) as that phenomenon in which "cases will group round the in- 
dividual perception as envisaging (without self-consciousness) that 
one immediate possibility of attainment which represents the closest 
analogy to its own immediate past." The interpersonal world of the 
submissive man tends to become quite lopsided, putting more and more 
pressure on him to obey and not to command. 

Survival anxiety presses the individual to repeat and narrow down 
his adjustive responses. He thus comes to a stable but restricted 
reciprocal relationship with his interpersonal world. But this is only 
one half of the total event. 

The Pressure to Change Responses. In addition to this tendency 
for the familiar personal environment to become limited in scope, we 
have seen that the environment at large presents one with a wide 
range of social stimuh. In any single day most individuals roaming 
around in their ecological space find suitable situations for expressing 
all sixteen interpersonal mechanisms. To the extent that the indi- 
vidual inhibits some of these, he is not employing the appropriate 
responses demanded by the environment. Failure to adapt to the world 
about it generally creates survival anxiety in the organism. 

The Insoluble Dilemma of AdjustxMent: Stability Versus 
Flexibility. The human being is, according to this view, caught be- 


tween two polar whirlpools of anxiety. Rigid repetition of inter- 
personal responses minimizes conflict and provides the security of con- 
tinuity and sameness — in Whitehead's useful terminology called 
"endurance." But the environment at large is not the same — and ad- 
justment to it demands a flexible generality of inteqjersonal response. 
The notational system and general premises of the present work lead 
us to conclude that this is the critical survival dilemma — the basic 
conflict, if you please, of human nature. A quotation from Egon 
Brunswik (3) appears pertinent here. He points out that "survival and 
its sub-units, which may be defined as the establishment of stable inter- 
relationships with the environment, are possible only if the organism 
is able to establish compensatory balance in the face of comparative 
chaos within the physical environment. Ambiguity of [stimulus] 
cues and means [i.e., organismic responses] relative to the vitally 
relevant objects and results must find its counterpart in an ambiguity 
and flexibility of the . . . mediating processes in the organism." ^ Or, 
to use Whitehead's words, ", . . every scheme for the analysis of na- 
ture has to face these two facts change and endurance^ The mo- 
mentum of the logic we are using in this book has led us to define two 
basic maladjustive factors in terms of these dichotomous sources of 
anxiety: rigidity, which brings a narrow adjustment to one aspect of 
the environment, and unstable oscillation which is an intense attempt 
to adjust to all aspects of the presented environment. These concepts 
involve the multilevel organization and diagnosis of personality and 
must await publication in a subsequent volume. 

Between the two maladjustive extremes of personality, rigid con- 
tinuity and oscillating noncontinuity, occur the greatest majority of 
human adjustments. Most individuals, as we have seen, tend to select 
a limited set of preferred reflexes which operate spontaneously, but 
not with inflexible repetition. The average individual is still able to 
call out automatically any and all reflexes along the continuum to meet 
the exigencies of the environment. In general orientation and in the 
crucial decisions of his life, he is likely, however, to have employed 
the narrowed responses. And he has very likely succeeded in training 
the significant "others" in his life to react in reciprocity to his inter- 
personal style. The average person has thus created himself and his 
world along the lines of a purposive but limited set of interpersonal 
relationships. He has worked out, usually by means of involuntary 
reflexes, a balance which is best calculated to meet the double threats 
of rigidity and chaotic flexibility. His Level I automatic communica- 
tions have provided him with smoothly operating techniques (^e- 

^ The italics and parenthetic notes are the author's. 


termining the nature of his self and world. Like his more neurotic 
brethren, he too gets from life the interpersonal returns for which he 
has bargained — just that and no more. 

The Principle of Reciprocal Interpersonal Relations 

The principle of self-determination as it operates at Level I has 
several implications. The notion that we must take the credit or blame 
for our own life situations has had an obvious effect on clinical prac- 
tice. It assigns to the individual patient the responsibility for develop- 
ing and managing his own personality. This is a terrible power that 
we assign to him, one which he is often not willing to believe or accept. 
The key factors in personality seem to be the purposive messages we 
express to others in our Level I communications. For many patients 
these are signals of weakness and blame: "Others must help me" and 
"Others are my undoing" are familiar and poignant themes expressed 
by many psychiatric patients. The notion of self-determination re- 
moves the protective devices of projection and externalization — giving 
in return a priceless, but often unwelcome gift of personal power. 

In developing these themes a rather curious imbalance may have 
been noted. For purposes of exposition we have concentrated on the 
viewpoint of the subject. At times it may have implied a paradoxical 
situation in which everyone goes around training others to respond to 
him in specified ways. This is, of course, rather puzzling. If everyone 
is actively creating his own interpersonal world, this leaves no one 
left to be passively trained by others. 

This dilemma is caused by the concentration on one side of the 
interpersonal exchange — the subject. Actually, we know that we can 
never understand interpersonal relationships unless we study both sides 
of the interaction. When we pause to isolate and study one side — the 
self or subject side — of interpersonal behavior, we do so at the risk of 
distortion. As we consider, in turn, the various levels and areas, we 
encounter the danger of segmental overemphasis — one of the plague? 
of psychological theory. The principle of self-determination is a 
probability statement which has reference to the global organization of 
personality in general and Level I in particular. The over-all system 
of the total personality is for all predictive purposes the unit upon 
which we focus. It has special importance in shaping a strategy and 
tactic of psychotherapy. It should be kept clear that in the preceding 
section we have, for expository purposes, stressed the "self" response 
and understressed the "other," or environmental factors. In actuality 
both partners in any relationship share the responsibility for its de- 
velopment — a mutual determining operation is occurring. The mother 
does not create the child's personality. The child does not create the 


maternal reaction. They both are engaged in a most intricate recipro- 
cal process to which both bring determinative motivations. 

Many Interpersonal Exchanges Reinforce the Original Re- 
flex. The time has now come to consider both sides of the interper- 
sonal situation — the two-person commerce of communication. The 
first point worth comment is the reinforcing quality of social interac- 
tion. Our actions toward other people generally have the effect of 
pulling a reciprocal response from them. This in turn tends to 
strengthen our original action. If you walk up and aggressively shove 
a stranger, the chances are good that he will shove you back. Of 
course, this rule does not work uniformly. One out of a hundred 
might be that Christian soul who would tenderly embrace you. A 
few might slink away from you. A few might docilely attempt to 
placate you. The largest percentage would mirror your aggression — 
and probably shove back. Your counterresponse then becomes the 
issue. You might apologize, you might retreat, but assuming you are 
an "aggressive shover" to begin with, the statistically probable re- 
sponse is to shove back, perhaps harder. 

You have provoked a response which has reinforced your original 
action. This reinforcing process has been dignified with the title of 
the principle of reciprocal interpersonal relations. This is a general 
probability principle. It holds that: Interpersonal reflexes tend {with 
a probability significantly greater than chance) to initiate or invite 
reciprocal interpersonal responses from the ^^other^^ person in the inter- 
action that lead to a repetition of the original reflex. 

Before considering the ramifications and quaUfications of this prin- 
ciple we shall glance at a few examples of its operation. 

How Group Therapy Patients Provoke Each Other to In- 
creasing Repetition. Group psychotherapy provides a splendid op- 
portunity to observe the development of interpersonal patterns. The 
members come together as strangers. The initial sessions are anxiety- 
provoking. The novelty and tension combine to produce interpersonal 
reflex behavior which is quite clear-cut. Unless a definite selection 
principle is employed, the members of a group tend to scatter around 
the interpersonal circle. That is, one or two will demonstrate helpless, 
dependent reflexes. One or two will be sullen and silent. One or two 
will be superior, mildly antagonistic. One or two will briskly begin 
to take leadership roles, try to "get the ball rolling," help the other 
patients, etc. The pattern of reciprocal reflexes that develops is rich 
and complex. 

For demonstration purposes it might be best to review a less hetero- 
geneous group. We think here of the group comprising three phobic, 


dependent women, two schizoid men, and a psychosomatic man. In 
the first session the three fearful women nervously described their 
symptoms and then fell into a protective silence. The schizoid men 
muttered their introductions and sank into an isolated retreat. The 
floor was left to the therapist and the psychosomatic man. 

The latter was a friendly, energetic, talkative person who rattled 
on for about ten minutes about his symptoms and his life situation. 
When he finished his competent and congenial narrative, the group 
fell into a prolonged silence. After two or three intensely long minutes 
of soundless hush the psychosomatic patient entered again with a ques- 
tion to the therapist. The patient then expanded on this topic for about 
five minutes. The tomblike silence resumed — interrupted only by the 
shifting of chairs and the rustle of smoking activity. 

The therapist then intervened to comment on the silence and asked 
each patient in turn what his associations were to the topic introduced 
by the psychosomatic patient. The therapist concluded the session 
by reviewing the silence and explaining that he had intervened at the 
end to help the patients learn how to communicate in the group. 

During the second meeting the same pattern repeated. Long silences 
developed. The tension clearly mounted during these lulls. The 
phobic ladies squirmed, looked uncomfortable, but kept silent. The 
schizoid men frowned, edged their chairs further toward the corners, 
and kept silent. The pressure on the friendly, talkative member would 
build up until he would finally begin to speak. He tried to get the 
others to talk. He asked them questions. He described at greater and 
greater length events from his own life (most of them concerning 
superficial events — hobbies, work experiences, etc.). By the fourth 
session the tension had mounted to an intense peak. The silences grew 
longer and more painful. The psychosomatic patient found himself 
involved in a series of monologues. It seemed that the patients were 
all getting disgusted with themselves and with each other, the one for 
talking too much and the others for not talking enough. By this point 
the talkative patient, in fact, found it hard to refrain from talking. 
When one of the others would venture a comment he would inter- 
rupt, ask questions, and relate his own associations. The executive 
outgoing patient had trained the others to be listeners. He later con- 
fessed that he rather fancied himself as a subleader in the group, and 
half-boastingly, half-sheepishly described his reflex skills in extro- 
verted glibness. 

The other group patients had successfully trained the extrovert to 
dominate them. They had forced him into a responsible, competent 
role and had thus emphasized and reinforced their own withdrawing 
tendencies. Their original reflex patterns had contributed to a tense 


situation. The more the tension developed, the more they increased 
their reflex techniques for handling anxiety. The psychosomatic pa- 
tient was almost frantically active and the others silent. They were all 
making a failure out of the group along the same lines of their life 
failures. When this reciprocal process had reached its optimal point, 
the therapist intervened to help the members to understand how they 
reacted to the tension and how they increased it by their reactions. 

We expect each group member to contribute to the failure of com- 
munication that tends to develop in our therapy groups. This initial 
breakdown caused by the reciprocal principle is allowed to develop in 
the early stages in the group since it allows each patient to repeat his 
interpersonal imbalances in the therapy situation. The very real ten- 
sion of the group situation provides valuable information about how 
each person handles anxiety. It is somewhat analogous to the trans- 
ference neurosis of individual therapy. It provides material for many 
months of subsequent analysis. 

This example centered around the reciprocal patterns of one mem- 
ber versus the group. More discrete interactions between pairs of in- 
dividual patients inevitably develop and pro\ide more complex and 
specific examples of the reciprocal process. Earlier in this chapter we 
have described a distrustful patient who was convinced that others 
were unsympathetic and mean to him. He provides another example 
of the reciprocal principle. We saw how this man easily and auto- 
matically provoked rejection and dislike by means of his growling 
suspiciousness. The hostile reception he received from the others led, 
of course, to an increase in his bitter distrust. He invited responses 
which led to a repetition of his original reflex pattern. 

Reciprocal Relations Are Probable, Not Inevitable. The re- 
inforcing process we have been describing is not an all-inclusive prin- 
ciple. It is a probability function. It does not necessarily hold for the 
individual interaction. Aggression usually breeds counteraggression. 
Smiles usually win smiles. Tears usually provoke sympathy. In spe- 
cific cases, however, these general rules break down. Aggression can 
win tolerant smiles. Tears can provoke curses. But, when we study 
the thousands of interactions that make up each day of social existence, 
the principle becomes increasingly useful. Many kinds of variation 
and inconsistency operate to lower perfect predictability of inter- 
personal behavior. The meaning of the cultural context, the personal- 
ity of the "other one," and oscillation tendencies in the individual are 
always complicating factors. Like any other principle which involves 
human emotions, the principle of reciprocal relations operates in 
probabilistic terms. 


Effect of the Other Person's Personality. Reciprocal rela- 
tions are more likely- to develop with certain personalities. The prin- 
ciple holds most uniformly with pairs of symbiotically "sick" people. 
A phobic, dependent wife and a nurturant, strong husband would be 
such a pair. The more the husband takes care of her, the more the de- 
pendence repeats. The more the wife clings, the more pressure on the 
husband to be gentle and protective. Even in a symbiotic marriage of 
this sort, the reciprocity would tend to break down if other motives 
enter the behavior of either. If hostile reproach lies behind the wife's 
weakness, or impatient superiority behind the husband's strength, then 
new chains of interaction may develop. 

Another aspect of this principle: The sicker you are the more 
power you have to determine the relationships you have with others. 
A maladjusted person with a crippled set of reflexes tends to over- 
develop a narrow range of one or two interpersonal responses. These 
are expressed intensely and often, whether appropriate to the situa- 
tion or not. Now a normal person has a fairly flexible range of re- 
flexes. He can use any interpersonal response if the situation calls it 
out. He is less committed to and, for that matter, less skillful in the 
use of any particular reflex. When the two interact, it is the "sick" 
person who determines the relationship. 

Suppose that the suspicious young man just cited meets up with a 
fairly well-rounded person. The latter may greet him cheerfully. The 
other may frown, or shoot a sharp glance, then cast his eyes to the 
ground. The normal person may invite the other to the movies — to 
which he replies with a sullen remark. No matter how flexible or well- 
meaning the one may be, the other will eventually force him to take a 
negative critical position. 

The more extreme and rigid the person, the greater his interper- 
sonal "pull" — the stronger his ability to shape the relationships with 
others. The withdrawn catatonic, the irretrievable criminal, the 
compulsively flirtatious charmer can inevitably provoke the expected 
response from a more well-balanced "other." 

The flexible person can pull a greater variety of responses from 
others — depending on his conscious or unconscious motives at the 
moment. He can get others to hke him, take care of him, obey him, 
lead him, envy him, etc. The "sick" person has a very narrow range of 
interpersonal tactics, but these are generally quite powerful in their 
effect. I have seen compulsive, responsible group members after sev- 
eral months of treatment desperately trying to get the other group 
members to understand and commiserate with their inner feelings of 
weakness and despair. They had trained them well to look up and 



respect them. Their own managerial reflexes kept firing even at the 
moment they were verbally appealing for help and sympathy. 

Variation Within the Individual Affects Reciprocal Rela- 
tionships. Another qualification of the principle of reciprocal rela- 
tions must be included. In describing human behavior the impression 
is often given that a consistent line of adjustment is exhibited. In most 
of the illustrations used in this chapter, the subject's role is made to ap- 
pear fixed. Actually, we know that inconsistency and changeability 
are the rule and not the exception in human emotions. The factors of 
change and stability will come under detailed survey in Chapter 13. 
They are, indeed, studied as a separate dimension of personality — the 
variability dimension. Included under this topic are all the measurable 
variations which affect human behavior — changes in cultural context, 
changes over time, changes due to conflict and variety among the 
levels of personality. 

At this point it is sufficient to point out that no interpersonal role 
is absolutely pure or rigid. The most withdrawn catatonic sends out 
occasional tendrils of affect. The most hardened criminal occasionally 
has a moment of congeniality. The most autocratic five-star general 
occasionally admits he is wrong. Most people show considerable con- 
flict or inconsistency in their actions from time to time. No matter 
how thick and effective the reflex defenses, underlying inconsistencies 
eventually manifest themselves. 

When this happens the principle of reciprocal relations tends to 
break down. The probable accuracy of the predictions drops. A 
flirtatious woman provokes seductive responses from a man. His ap- 
proaches set off stronger flirtatious actions. The man becomes more 
seductive. At some point in this process underlying motives may step 
in to change the pattern. In some cases, a flirtatious fa9ade may cover 
deeper feelings of competition or contempt toward men. The woman 
would then shift to behavior which Erickson describes as "bitchy," 
and rejecting. The reciprocal pattern of entice versus seduce would 
shift. The man's reaction would then vary depending on the nature 
of his multilevel pattern. He might continue to seduce, he might be 
hurt, he might become dependent. 

The same process of circular interactions leading up to an intense 
breaking point often occurs between parent and child. Dependence 
pulls nurturance which provokes further dependence — . In some 
cases the spiraling increase in intensity leads to a temporary crash. At 
some point the parent's underlying feelings of selfishness or self-pro- 
tection lead to refusal. Father comes home one night tired and 


grumpy. Outside events may have set off underlying feelings of dep- 
rivation, or self-pity, or sadism. He may snarl at the child. The child 
then whines. The whining might increase the father's irritation. A 
new series of reciprocal events may thus be initiated. 

Alternation of behavior is, of course, not an unhealthy manifesta- 
tion. Moods shift; we carry over the feelings from one situation into 
another. Events of the day set off underlying effects which may be 
quite different from the current reality situation. It is safe to suggest 
that everyone acts inappropriately many times each day. These incon- 
sistencies can hardly be considered abnormal. The lines of inter- 
personal communication are constantly breaking down momentarily, 
but these involve no permanent disasters. A healthy father-child re- 
lationship is not paralyzed because one of the two has a "bad day" or 
carries over inappropriate effects. 

On the other hand, very rigidly formed relationships can be upset 
badly by shifts in the pattern of reciprocal relations. Some institu- 
tional relationships are very inflexible and demand perfect reciprocity. 
The army officer expects to provoke consistent obedience. A rent in 
this kind of interpersonal fabric can be seen as unforgivable. Some 
kinds of symbiotic marriages are so rigid that deviation in reciprocal 
roles can cause intense anxiety. When a servile, docile husband shows 
a flash of rebellion against a dominating wife, the results can be ex- 

Thus, we see that many factors tend to qualify the principle of 
reciprocal relations. Among these we have considered variations in 
the cultural context, variations in the personality of the "other one," 
and variations due to multilevel ambivalences in the subject's personal- 

Multilevel Reciprocity Patterns. We have very little system- 
atic knowledge about interpersonal relations. We do know that a 
most complex, shifting matrix of forces operates in the simplest inter- 
action. Throughout this book we are forced to limit the theory, the 
illustrations, and the measurements to the simplest forms of interac- 
tion. One example of a rich and vital phenomenon which is at present 
beyond reach of our system has to do with multilevel reciprocal pat- 
terns. Complex patterns of interaction exist at all the levels of per- 
sonality. In some cases the smooth flowing exchanges of one level are 
threatened and destroyed by clashes caused by underlying variations. 
That is, two people may interact in a most automatic and rewarding 
pattern at the level of the interpersonal reflex. The seductive man and 
the flirtatious woman is one such situation. At the private level the 
feelings of both partners may be quite different. The man may have 



000 S'SSS- S!='55~ 

a o 00000 

lEuijoiuadXH Q 



OrrsOo >Ot1- — \o (^iO»viirs OOOO'*- 




O O") r.-, — . >0 "r^ 

Q h ^ '-. r-i '■' (^1 vo r- t-~ ^O r<-, r^ r^ ./^ TA O r-4 — ' O 

^ -SUIDEi^a-JpS a:: " - - ^ rv. 

g I OUSipeS ttj „ ,, o ^ O rv, .^ ^. ^ 00 r^. O r. O .^. -. O 

g h. -3AISS3J33V Q "^ "^ " -' -' 

tJ ^ OUSISSpjB^I Ov.0 000 Or-1-.-f O^fMOO r^--^-^0 

2 """ -3Ann3duio3 * ~' ~' "■ n — — 

>^ i-v, 1^, rn — I — . r^ _■ <> 

o ^ 

w 4J ii 2 -7; 1J' 


^ ^ S S = 

j3 - J^ c (u 

C -S Qj •- '*^ 

IS ^ 

1 "O 

lilts i^ii i«=si §1 .^1 

I .2 « e 0. >>^ -i I, -5 '^ e « '2 '5 -f 2 2 

S ^^ <=^ >-Sc>:. "2;^ >^ 4j c c o o (£ 


deeper sadistic motives toward women. The flirtatious girl may have 
underlying needs to reject and humiliate men. What starts off as a 
most spontaneous and fluid friendship eventually ends in a brawl. 
The underlying feelings of the participants do not lend themselves 
to a durable relationship. 

This aggressive ending might not occur if the underlying feelings of 
the partners were reciprocal. Suppose that the woman's underlying 
feelings, instead of being competitive and rejecting, were masochistic. 
Their surface reflexes blend nicely into the pattern of seduce and en- 
tice. Their deeper feelings would, in this case, also blend nicely into 
the reciprocal pattern of sadism-masochism. The man's preconscious 
aggressiveness would tailor nicely into the woman's private needs to 
be aggressed upon. It is possible that many symbiotic marriages exist 
in which multilevel needs of both mates fit together into multilevel 
patterns of reciprocity. Our measurement methods are far from being 
able to tap these intricate networks which seem to characterize even 
the simplest relationships. 

Incidence of Level I-M Behavior in Various Cultural Sam- 
ples. A summary of the research findings concerning Level I be- 
havior is presented after each of the eight clinical chapters (Chapters 
15-22). At this point, to give an overview, it may be helpful to list 
the percentage of Level I-M types found in several institutional or 
symptomatic samples. 

In the preceding chapter a method was described for summarizing 
interpersonal behavior at any level in terms of a single point on the 
diagnostic grid. The location of this summary point determines the 
interpersonal diagnosis (see Chapter 12). Thus if the resultant of the 
Level I scores for an individual locates in the AP octant he is diag- 
nosed as a Managerial-Autocratic personality at this level. Table 2 
presents the percentage of cases in fifteen samples falling in each 
octant at Level I-M. 

A detailed re-examination of these data will be found in later chap- 
ters, but some of the meaning of Level I-M behavior can be derived 
by inspection of the table. Some samples (military officers, normals, 
psychosomatic cases) emphasize strong, hypernormal facades (octants 
AP and NO). Those samples which include people in trouble (prison- 
ers, psychiatric patients) manifest more alienated or passive behavior 
(octants DE, FG, or HI). 


1. Bales, R. F. Interaction process analysis. Cambridge, Mass.: Addison-Wesley 
Press, 1950. 

2. BiON, W. R. Experience in groups: III. Human Relations, 1949, 11, No. 1, 13-22. 


3. Brunswik, E. The conceptual framework of psychology. International encyclo- 
pedia of unified science, Vol. /, No. 10, Chicago: University of Chicago Press, 
1952. Copyright 1952 by The University of Chicago. 

4. Cassirer, E. An essay on man. New Haven: Yale University Press, 1944. 

5. Lawrence, D. H. Women in love. Modern Library ed. New York: Random 
House, Inc., 1920. 

6. Leary, T., and H. Coffey. The prediction of interpersonal behavior in group 
psychotherapy. Psychodrama and gr. psychother. Monogr., 1955, No. 28. 

7. Mead, G. H. Mind, self and society. Chicago: University of Chicago Press, 1934. 
Copyright 1934 by The University of Chicago. 

8. Powelson, D. H., and R. Bendix. Psychiatry in prison. Fsychiat., 1951, 14, 73-86. 

9. Sapir, E. Speech as a personality trait. Amer. J. Sociol., 1927, 32, 892-905, Uni- 
versity of Chicago Press. Copyright 1927 by The University of Chicago. 

10. Sw^EET, Blanche. A study of insight: its operational definition and its relation- 
ship to psychological health. Unpublished doctor's dissertation, University of 
California, Berkeley, 1953. 

11. Thelen, H. a. Method of sequential analysis of group process. Mimeographed 
working instructions, 1952. 

12. Whitehead, A. N. Science and the modem world. New York: Macmillan, 1925. 


The Level of Conscious Communication: 
The Interpersonal Trait 

This chapter is devoted to an examination of Level II, the data of con- 
scious description. We deal here with the individual's perceptions of 
himself and his world as he reports them. 

Like the other levels of personality, this one is automatically defined 
by the data which contribute to it. There is only one criterion for de- 
termining Level II data: conscious verbal report by the subject. We 
are interested in ivhat the subject says, the content of his verbal ex- 
pressions. From these we focus on the interpersonal themes which he 
attributes to himself and to "others." From these we obtain the 
variables of Level 11. 

It must be noted that the consensual accuracy or truth of these 
verbal reports has no bearing on the definition of the level. If the sub- 
ject says he is popular, the appropriate Level II code for this inter- 
personal role is assigned. Now dozens of observers may agree that 
he is quite unpopular with his associates. This fact shows up on our 
diagrams for Level I. But the Level II rating concerns not what he 
does, not what he privately thinks or wishes, but what he says. 

This is called the level of conscious description because it reflects 
how the subject chooses to present himself and his view of the world. 
It will be noted that we do not call it the level of consciousness, but of 
conscious communication. This is an important distinction. The 
phenomenon of consciousness is one of the most elusive issues in the 
history of Western thought. One of its most confusing aspects is, of 
course, its subjective nature. The scientist can never understand or 
measure what another person has in his consciousness. It is often quite 
difficult for the subject himself to know the focus and limits or his 
awareness. Between the subject and the psychologist there exists any 
number of potentially distorting factors — deliberate omissions, expres- 



sive inaccuracies, and the like. And we never know the exact level of 
awareness from which the statements come. 

Since it is impossible to obtain an objective evaluation of the sub- 
jective viewpoint of another person, many psychologists have at- 
tempted to discard the whole issue of consciousness. But in so doing 
an essential dimension of human behavior is lost. 

Two principles must be applied to any scientific approach to the 
conscious aspects of pei-sonality. The first is the classic solution de- 
veloped (but not utilized) by the earliest behaviorists: treat the sub- 
ject's introspection not as the essence of truth, but as a behavioral ex- 
pression to be evaluated in the light of all the other measurements. 
The second principle is an explicit corollary that can only develop 
from a systematic multilevel analysis of behavior. It holds that the 
data of conscious report have of themselves an ambiguous meaning 
until they are systematically evaluated in the light of the data from 
the other levels of behavior. 

At Level II we deal, therefore, with conscious reports and not con- 
sciousness. We define it operationally in terms of all the statements 
an individual makes about himself or his world. We employ it and 
evaluate it in relation to other levels of personality. 

The Attributive Nature of Personality Language 

Transcriptions of everything that a patient says during an hour of 
psychotherapy provide one source of raw data for Level II measure- 
ments. The patient's testimony in this form can then be studied from 
the standpoint of the interpersonal system. Everything that the patient 
says about himself becomes Level II "Self." Everything he says about 
the people in his interpersonal world becomes Level II "Other." The 
accuracy, the deeper significance, the immediate purpose behind these 
responses is disregarded. Their direct surface meaning is the essence 
of the Level II classification. 

In studying the verbal content of these descriptions an interesting 
fact develops. They are all attributive or adjectival. They are all 
significations. They can all be interpreted as assigning a quality to the 
self or the world. They can, thus, be reduced for analysis to a descrip- 
tive adjective or to adjectival phrases. The interpersonal context of 
everything that is said about oneself or one's world can be translated 
into a generic attributive form: "I am a person, in relation- 
ship to 

Take, for example, the patient's testimony, "I was really angry at 
my boss today. I took it docilely for a while. Finally I insulted him. 
I've always hated his guts." The subject and object of this inter- 
personal relationship are clearly self-boss. The four sentences vary 


in their mode of expressing the interpersonal theme — adjective, verb, 
participle — but they can all be translated into the attributive formula: 

angry -| 

I feel like an: insulrine- I P^^^^" ^^ relationship to my boss, 
hating J 

At the level of conscious description we deal with the subject's 
language about himself and others. The interpersonal attributes are 
the specific rated units. The interpersonal themes expressed in these 
significations are coded according to the matrix of sixteen variables 
and provide a systematic summary of the subject's view of himself and 
his world. The operational procedure for defining a level in terms of 
the source of the language is not unique to this level. It is now pos- 
sible to look back at the Level I reflex communications and see that 
they are also defined by the source of the language — the attributive sig- 
nifications of the observers who rate the subject's behavior. At Level 
I we do not deal with the reflex conversation of gestures itself but 
measure its effect on others. We ask the individuals who observe or 
interact with the subject to make attributive statements about him 
which reflect his social stimulus value. The language by which the 
subject is described, by others or by himself, comprises the data for 
Levels I and II respectively. 

The data from Level I can therefore be translated into the same 

type of atributive formula. ''He is a person in relation 

to This systematic^ approach to the linguistics of person- 
ality provides a direct method for comparing the levels of personality. 
The relationships between levels — discrepancies, concordances — de- 
fine another dimension of personality, the variability dimension. The 
direct measurement of these mechanisms which thus relate the levels 
of personality is made possible by the rigorous analysis of the lan- 
guage of personality. 

The Measurement of Interpersonal Attributes 

The unit with which we measure the language of conscious descrip- 
tion is called the interpersonal attribute or the interpersonal trait. 
These terms have been selected because they reflect the adjectival or 
attributive nature of the Level II data. We classify Level I behavior 
in terms of interpersonal reflexes, gestures, or mechanisms. We clas- 
sify Level II behavior in terms of the interpersonal attribute or trait. 

The interpersonal trait of Level II is formally defined as the inter- 
personal motive attributed by the subject to himself or another in his 



conscious reports. Every discernible or ratable interpersonal theme 
in the content of the individual's verbalizations defines a unit of Level 
II behavior. 

The themes or categories employed are derived from the circular 
continuum of interpersonal variables. It w^ill be recalled that in meas- 
uring the subject's reflex behavior an inexhaustible list of sample verbs 
was held to apply to each of the sixteen generic interpersonal pur- 
poses. The same procedure is followed for measuring attributive be- 
havior of Level II. The interpersonal traits were developed by simply 








^\A p 


1 — ~~~--~.^ 

'^i! ^-C ' 












E T 

? ^ 






1 > 

J"^e■'s^■ /CAN BE 
'^'^ / OBEDIENT 



^fslSe7^f- USUALLY 











Figure 6. Interpersonal Check List Illustrating the Classification of Interpersonal 
Behaviors into Sixteen Variable Categories. 


replacing the verbs of Level I with the coordinate or appropriate ad- 
jective. The adjectives which go with the reflex (i.e., verb) to com- 
plain would obviously include complaining, resentful, bitter, etc. The 
adjectives which parallel the reflex to love would be loving, affection- 
ate, etc. This translation of verbs into adjective equivalents is not al- 
ways so Hnguistically simple. There are many interpersonal reflexes 
for which equivalent adjectives do not exist. Extensive methodological 
procedures have been carried out in order to deal with these technical 
difficulties. As a result of these exploratory studies, the interpersonal 
meaning of most words in the English language which have a social 
connotation has been determined in terms of the sixteen-point con- 

Several adjectives characteristic of each generic interpersonal trait 
(i.e., each point on the circular continuum) are included in Figure 6. 
These adjectives are suggestive and illustrative. They are by no means 
exhaustive of the entire range of traits which fit each point of the 
circle. In analyzing the traits employed by an individual, we rate not 
only the kind but the intensity of each attribute. Extreme, inap- 
propriate, and maladjustive interpersonal behaviors are thus distin- 
guished from the moderate and appropriate. The general nature of 
these adjustive and maladjustive traits is suggested by the words listed 
in the inner and outer rings respectively in Figure 6. 

Four Methods for Measuring Level U Behavior 

In Chapter 6 it was pointed out that several methods exist for ob- 
taining the data for any level. Whenever Level II data are being dis- 
cussed it is necessary to indicate the specific source — that is the opera- 
tions through which the data were derived. 

When trained personnel rate the verbal content of diagnostic inter- 
views, i.e., the patient's descriptions of himself and others, the result- 
ing data are assigned to Level Il-Di. These ratings are made from on- 
the-spot observations, notes, recordings, or transcriptions. 

When trained raters judge the verbal content of therapy interviews 
(group or individual) the descriptions of self and others are coded 
Level II-Ti. 

Scores from the Interpersonal Adjective Check List on which the 
patient rates his view of self and others are coded Level II-C. 

Ratings by trained personnel of the conscious descriptions of self 
and others taken from autobiographical essays written by subjects are 
coded Level II- A. 

There are, then, several methods for obtaining Level II data. The 
essence of them all is that we get the subject to describe himself and 
others. These reported perceptions are then scored in terms of the 


circular continuum. If the data are obtained through fixed, prepared 
test stimuh- questionnaires, check Hsts, and the hke, predetermined 
ratings assigned to each test item make the scoring automatic. If the 
data come from free responses — conversations, interviews, autobiog- 
raphies — then two or more trained technicians independently rate 
each interpersonal reference. 

The rating of interpersonal attributes at the level of conscious 
description is illustrated in the following examples: 

The scoring of interpersonal traits at the conscious level. To illustrate the 
rating of interpersonal traits at the conscious level, there follow examples of 
ratings of an adjective check list, the content of therapy sessions, and an auto- 

a. The Scoring of Interpersonal Traits as Applied to an Adjective Check 
List, Level II-C: 

Adjective Trait Intensity 







b. The Scoring of Interpersonal Traits as Applied to the Content of Dis- 
cussion of a Group Psychotherapy Session, Level II-Ti: 

Other Self 

And since I've been married I've been able to make a sub- 
stitute, a transference of these feelings from my mother to 
my wife. I guess I depend on my wife a lot, more than I K-3 

C-3, A-2 should. She is a strong person. She admires strength. I 
think it makes her retract, withdraw from me when I am 
dependent on her. This makes me feel worse. Then she K-3 

C-3, A-2 tries to drive me. She's an ambitious person. It makes me 

feel very helpless. This is a real vicious circle and it has me 1-3 

c. The Scoring of Interpersonal Traits as Applied to an Autobiography, 
Level II- A: 

Other Self 

I guess I was a very co-operative child, but this is just from L-3 
what I've heard. I've always been timid all my life. This is 7-3 
especially true with girls. I believe my parents realized this 

A-2 and often, especially my mother, tried to push me a little 

which I resented and probably went out of my way to do B-3 
the opposite. I have always resented and still do, being told F-3 

A-'i what to do or obviously being led. 








The simplest and most standardized method for obtaining an esti- 
mate of Level II perceptions is to employ the Interpersonal Adjective 
Check List. This test has been specifically designed by Robert Suczek, 
Rolfe La Forge, and others to fit the matrix of the sixteen interpersonal 
variables. In its present form (Form IV) it consists of 128 adjectives, 
8 for each point on the circle (see Appendix 2). The check list is 
calibrated in four degrees of intensity and the array of adjectives is 
balanced according to the expected frequency of usage. The patient 
simply checks all of the items which he believes describe his behavior. 
Since each term is already prescored, his responses automatically 
produce his Level II-C self-pattern.^ The patient can be asked to use 
the same check list to rate the significant "others" in his life. This 
gives a standardized picture of his own description of self and world. 

The Patient Diagnoses Himself 

Let us consider some illustrations of Level II measures from a clinic 
patient with a history of chronic maladjustment. The data were col- 
lected from the Interpersonal Adjective Check List Form IV on which 
the subject (at the time of entrance to the psychiatric clinic) succes- 
sively rated himself and his family members. 

Figure 7. Level II-C Self-Description of Illustrative Subject 
Based on Interpersonal Check List. Key: Radius of circle equals 
16 check list words. 

Figure 7 presents a diagrammatic summary of his self-descriptions. 
It is clear that the patient sees himself as an "unbalanced" person. He 
has consistently checked himself as being distrustful, passively hostile, 
and isolated. The marked imbalance indicates that he claims these 

* A table of norms for converting Level II-C dominance and hostility indices into 
standard scores will be found in Appendix 5. 



traits to the extreme degree, thereby diagnosing himself as disturbed 
in his interpersonal behavior. Turning our attention to the inter- 
personal themes that he does not attribute to himself (the blank area 
of the circle) we see that he clearly denies all the affiliative (X, L, M, 
N, O) and strong assertive (P, A,B,C) feelings. 

We have here the patient's self-diagnosis — a most important shce 
of the entire personality pattern. The Level II-C self-profile has 
considerable clinical significance. Several probabiUty laws hold for 
this single measure. 

A Patient Diagnoses His Family Members 

The patient's descriptions of his family members provide another 
set of valuable data. Considering his view of his father (Figure 8), two 
statements are immediately pertinent: (1) he diagnoses his father as a 
distrustful (G), passively hostile (F), and isolated (H) person; (2) he 
sees his father as being very much like himself. It is important to note 

Figure 8. Level II-C Conscious Description of Father by 
Illustrative Subject Based on Interpersonal Check List. Key: 
Radius of circle equals 16 check list w^ords. 

the difference between these two statements. To borrow the vocab- 
ulary of the logician, the first is a class statement about a single area of 
personality, his view of his father. The second is a relationship state- 
ment comparing two discrete areas of personality — Level II-C "Self" 
versus Level II-C "Father." 

Turning to this patient's view of his mother (Figure 9) we observe 
a contrast. The mother is seen as unyielding (B), rejecting (C), and 
punitive (D) to an extreme degree. He does not attribute any affec- 
tionate or passive qualities to her. Comparing the view of mother with 


his own self-perception (shifting thereby from a class to a relationship 
context of discourse), a marked discrepancy becomes apparent. He 
sees himself as being like his father but unlike his mother. 

Figure 9. Conscious Description of Mother by Illustrative Sub- 
ject Based on Interpersonal Adjective Check List. Key: Radius of 
circle equals 16 check list words. 

Level II Provides a Measure of Conscious Identification 

By inspection or by quantitative comparison we can determine the 
similarity-difference factors relating the self-profile to the Mother and 
Father circles at Level IL These relationships comprise the network of 
measurable phenomena called variability indices, which serve the func- 
tion of relating the areas and levels of personahty. We have suggested 
here that the relationships between Level II Self and Level II Other 
can be called identification or disidentification. To illustrate some of 
these relationship mechanisms as they operate at this level of personal- 
ity we shall construct a diagrammatic summary of this patient's Level 
II perceptions (see Figure 10). The lines linking the summary points 
provide a linear index of the arithmetic discrepancies between the areas 
of personality involved. The longer the line, the greater the difference 
in interpersonal themes attributed to the persons in question. It is 
then possible to translate this diagram into a verbal summary of the 
Level II behavior, employing the useful, but semantically suspect lan- 
guage of the clinic. 

It might be said that this patient sees himself as exploited and re- 
jected in relationship with an unsympathetic and cold mother. He is 
consciously identified with a weak and distrustful father. He is con- 
sciously disidentified with his mother. 



The patient's view of his wife (see Figure 10) adds another factor 
to the picture. He tells us that she is a hard-hearted, hostile, and re- 
jecting person. He consciously equates his wife with his mother, at- 



Figure 10. Diagnostic Summary Profile of Level II Self and Other Scores for 
Illustrative Patient. Key: The center of the diagnostic circle is determined by the 
mean of a clinic sample of 800 cases. The placement of the three summary scores 
is determined by trigonometric formulas (see Chapter 6) which yield horizontal and 
vertical indices. These are converted to standard scores. The lines between the sum- 
mary points provide linear estimations of the amount of conscious identification or dis- 
identification (see Chapters 6 and 13). 

tributing the same interpersonal motives to both. A wealth of clini- 
cal cues is summarized in the family descriptions. From them we ob- 
tain the patient's conscious diagnosis of his own oedipal situation, his 
marital relationship, and his relationships with three central figures in 
his life. Many probability laws hold for each of these measures. His 


view of parents is correlated with psychiatric diagnosis, symptom, 
and with the intensity and type of underlying conflict. 

In one sense these systematic measurements are quite limited. A 
clinical interview would give the same data just as easily and would 
provide a much more rich, specific, and sensitive registry of these facts. 

The circular profiles have some compensating virtues — they are 
reliable, they are quantitative, and they are standardized and cali- 
brated in terms of the sixteen variables by which we measure inter- 
personal behavior at other levels. This means that we can directly 
compare the different perceptions which the patient reports. They 
allow us to build up a series of probability laws which hold for each 
level and for the relationships among levels. 

Level II Presents the Patient's View of the 
'^Transference'' Situation 
We have reviewed how a patient entering the Kaiser Foundation 
Psychiatric Clinic diagnoses himself and his family members. The 
relationships among these measures have provided indices of conscious 
identification and conscious equation. When the intake evaluation 
was completed, this patient began psychotherapy. After nine hours 
of treatment he filled out the Interpersonal Check List on his therapist. 
This gives us his conscious description of the therapist (Figure 11). 
The patient diagnoses the doctor. 

Figure 11. Level II-C Description of Therapist by Patient 
Illustrating a Measure of (Conscious) Transference. Key: Radius 
of circle equals 16 check list words. 

This patient reported his therapist (Figure 11) as a well-balanced 
person. He did not use intense or one-sided descriptive terms. He 
attributed moderate themes of strength (A), punitive firmness (D), 


and mild rejection (C) to his therapist. A mild negative transference 
at the conscious level is apparent. 

When the view of therapist is compared with his pretreatment view 
of self, we see that the patient is disidentified with the therapist. He 
reports his therapist as being much more like his mother and his wife 
(Figure 12). 

1 bitter Mother 1 cold 

Self I and ^-^ Wife I and 

J depressed Therapist J punitive 

In this manner the patient gives us a systematic picture of his oedi- 
pal and his transference situations— fro?;? his oivn viewpoint. The pa- 
tient's reports about these relationships may be quite different from 
the therapist's. The latter might not consider himself as being cold 
and strict with the patient. If the patient's description of his therapist 
is consensually inaccurate, this fact takes on a considerable impor- 
tance in understanding the treatment relationship. The relation be- 
tween the patient's view of another and the consensual view of that 
person allows for an operational definition of a classic defense mech- 
anism — projection. Where this inaccurate perception involves the 
therapist, we have obtained a measure of transference-projection. 

Therapists Can Measure Their Own Misperceptions 

The Kaiser Foundation research project has undertaken extensive 
studies of process in psychotherapy. The aim of these studies is to 
apply the interpersonal system to the therapeutic interaction and to 
the perceptions of the patient and therapist. The working principle 
employed in these studies is: the patient and therapist comprise a basic 
interacting unit. We do not study the patient in therapy, but both 
the patient and the therapist as they interacted. These therapeutic 
studies cannot be included in this diagnostic monograph, but they are 
worth brief comment here because they illustrate the application of 
Level II measurements. 

One procedure commonly employed is to have the therapist fill out 
an Interpersonal Check List on his patient. This gives us a most in- 
teresting measure. It tells us how the therapist sees the patient with 
whom he is in relationship. Figure 1 2 presents the therapist's picture 
of the patient we have been discussing. The circle tells us that the 
therapist sees the patient as deferent (/) and dependent (/). Now 
this description may or may not be consensually accurate. Regard- 
less of its "pull," it does summarize some valuable information — it tells 
us something about the conscious countertransference. 


We recall that the patient described the therapist as cold and re- 
jecting. This implies a certain fear and passive hostility on the part 
of the patient. He feels mildly rejected. The therapist sees the pa- 
tient as mainly weak and dependent. This suggests certain discrepan- 
cies in the communication pattern between the two. The therapist 
might fail to sense the- patient's feelings of deprivation, and assume 
deference and collaboration. 

Figure 12. Level II-C Description of Patient by His Psycho- 
therapist Illustrating a Measure of (Conscious) Countertransfer- 
ence. Key: Radius of circle equals 16 check list words. 

Fitting together the reciprocal perceptions by both members of the 
relationship often reveals striking breakdowns in communication. 
Projection and perceptual distortion on the part of patient and thera- 
pist often become apparent by the use of Level II measures. - 

This discussion brings us to a tricky problem of definition. We 
have been talking about Level II behavior — the conscious descriptions 
of self-and-other by the patient. When we introduced the therapist's 
view of the patient we complicated the issues. From the standpoint of 
the patient, the therapist's view-of-patient is Level I. The therapist 
is thus an outsider rating the patient. 

But in any study of therapeutic interaction we focus equally on 
patient and therapist. The therapist's view of the patient is a Level 
II Other measure from the standpoint of the therapist. Studies in 
interpersonal relations which attempt to use multilevel patterns of 
response can become quite complex since we must study both sides of 

^ A most ingenious research which illustrates the phenomenon of countertrans- 
ference, as measured by the interpersonal system, has been completed by Richard 
Cutler (1). This research deals with misperception of self and others in the psycho- 
therapeutic simation. 


the transaction. A's report of B is subjective (Level II) from the 
standpoint of A, but it is objective (Level I) from the standpoint of B. 

The specific definition, conceptuaHzation, and clinical meaning of 
these relationship variables (e.g., identification, projection) need not 
be taken up in detail here. In the context of this discussion of Level II 
behavior it need only be suggested that the relationships between self- 
perception and perception of "others" have considerable importance. 
Lawful connections do exist between these self-and-other circles. 
Patients who describe themselves as distrustful and isolated tend to 
present predictable pictures of their parents and the significant "oth- 
ers" in their lives. So do the patients who assign themselves to other 
extreme positions on the Level II circle. 

A host of low-order predictive functions can be called into play 
if we obtain this one type of personality measure — a patient's con- 
scious description of self. It should be kept clear that these are not 
foolproof prognostications. They are probability statements which 
allow us to make such predictions as, "If the patient describes himself 
as siveet and docile (/K), the chances are 5 to 1 he will attribute to 
at least one parent ideahzed, tender nurturance, and the chances are 
2 to 1 that he will see both parents in this way." From Level II Self- 
description alone we obtain a large but loose network of low-order 
probability statements which make predictions about other levels and 
areas. This is interesting theoretically, but of restricted practical value. 
When we add the data from another level or area — his view of parents, 
for example — the additional evidence tightens up the network of re- 
lationships. It increases the complexity of the personality structure in 
a geometrical rather than an arithmetical proportion. That is, it multi- 
plies the permutations and combinations of relationships. It also in- 
creases the accuracy of prediction. 

Use of Level II Patterns in Child Guidance 

We have stressed the point that Level II conscious descriptions are 
the most simple, straightforward measures of personality and their 
maximum usefulness is found in combining or comparing these con- 
scious reports with other levels. 

In child guidance. Level II patterns seem to have an especially 
valuable application. Diagnostic evaluation of children's cases is a 
complex process. One of the difficult aspects of this procedure is 
caused by the multiplicity of interpersonal relations involved. Under- 
standing the child's situation requires some knowledge of the child, his 
siblings, both parents. Often the parents' attitudes toward child-rear- 
ing are closely related to their own parents. Thus, a three-generation 
matrix of relationships can be involved. 

Mother's view of: 






Figure 13. Familial Pattern of Interpersonal 

Father's view of: 




S (H . 


Relations for Child Guidance Evaluation. 



A convenient way to systematize these patterns is to obtain Level 
II-C descriptions from both parents. The mother, for example, is 
asked to rate herself, her husband, her child, and both of her parents. 
The husband does the same. As indicated in Figure 1 3 a matrix of ten 
ratings is obtained. This three-generation pattern of interpersonal 
scores provides a large number of cues for understanding the child 
and his emotional background. 

The family constellation diagrammed in Figure 1 3 reveals that the 
mother sees herself as responsible, generous, and hypernormal; that she 
sees her child as rebellious and passive-resistant, her husband as being 
like the child. She describes her mother as being a strong, responsible 
figure and her father as being a bitter and defeated man. The father, 
on the other hand, describes the child as being less rebellious — which 
suggests that most of the friction (and perhaps the motivating force 
in bringing the child to the clinic) centers around the mother. The 
father tends to see himself like the child and like his own father. He 
describes his wife as being very much like his mother. A fairly wide 
discrepancy exists between the wife's self-description and her hus- 
band's view of her — he attributes much more hostility and bossiness 
to her than she admits. This interwoven pattern of mutual mispercep- 
tion and oedipal themes suggests several hypotheses which may be 
useful in understanding this family's situation and the problems which 
led the parents to bring the child to the clinic. 

According to Mary Sarvis (2), it is not unusual to find that the par- 
ents have quite different perceptions of the child, or that a certain 
personality formation characteristic of a grandparent has skipped a 
generation and reappeared in the child. 

When these scores are compared with each other, a complex net- 
work of discrepancy indices can be derived. We can measure the dif- 
ference between the husband's view of himself and his wife's percep- 
tion of him; or the similarity between the wife's view of her father, 
her own husband, and the child. These relationships will be given 
operational definition and further theoretical consideration in Chapter 
13, which deals with the measurement of variability indices. 

Variability and the Sublevels of Consciousness 

At the end of the last chapter, it became necessary to consider the 
objection that our clinical measurements are infinitesimally narrow 
contrasted with the broad variety of human behavior. In each differ- 
ent situation and at every point in time we deal with a changing or- 
ganism. Our measurements, however elaborate, are generally limited 
to a restricted range and to a fleeting span of time. It has been sug- 
gested that the predictions be limited to the context in which the 



measurements are made. The Kaiser Foundation system is a func- 
tional clinical system. By aiming our predictions at the further be- 
havior of the patient in the clinic, we use our information in the same 
context in which it was collected. 

Now, as we come to the close of this chapter, we are faced again 
with the same issue. The range of conscious reports is diverse. A per- 
son describes himself in a variety of ways, depending on his purposes 
and the environmental, situation. He will emphasize certain trends 
when he attempts to impress, others when he attempts to excuse him- 
self, others when motivated to confide. How do we know that the 
measurements of Level II we make in one or even several moments of 
time reflect the over-all scope of the individual's conscious percep- 
tions? Of course, we don't know. Here we must resign ourselves to 
the familiar indeterminism. We can never hope to sample the breadth 
of the individual's self-descriptions as they vary in time and context. 

There are, fortunately, several steps which can be taken to limit our 
ignorance. The first of these derives from the organismic premise that 
no datum of personality can be evaluated except in the context of the 
total organization. The tenor of our conscious reports is, as we well 
know, related to the Level I situation. What one says depends upon 
what one purposes. It depends upon the pressure of the social en- 
vironment. Let us illustrate. Consider an initial treatment interview 
in which the therapist is reflexly and unconsciously pushing the pa- 
tient to free-associate and confide. Let us assume that the interview 
has been recorded and the independent judge rates this behavior as 
directive (AP at Level I Other) behavior. Let the patient be B (stub- 
born, resistive) . We shall skip the question of responsibility for initiat- 
ing the relationship, i.e., who provoked whom to develop these roles. 
In most relationships this is a mutual process of training each other. 
In this context, the therapist might make a brief didactic remark to the 
effect that people sometimes have feelings about their childhood or 
about their parents that are important. The patient might produce the 
conscious description of self and other that she has nothing but the 
most loving feelings toward her parents who have always been kind 
and good to her. The following oversimplified formula has developed: 

Self (Patient) Other 

Level I B < — > AP (therapist) 

Level II M < — > O (parents) 

We verbally summarize by saying: "When the patient is being 
resistive or defensive to a directive therapist, she reports herself as af- 
fectionate to her idealized parents. Let us go on to assume that the 


therapist works through this power struggle and that by the twentieth 
treatment interview a participant phase of the relationship is develop- 
ing. The therapist is communicating support (N), and the patient is 
attempting to be cooperative (L), In this context she might confide, 
"Many times I have been disappointed and hurt by my parents' un- 
willingness to understand me and my point of view," The formula 
for this sequence becomes: 

Self Other 

Level I L < — > N (therapist) 

Level II G < — > C (parents) 

This translates back to the verbal summary: When the patient is 
confiding and cooperative to a supportive therapist, she describes her- 
self as rejected by her parents who are seen as unsympathetic. 

Here the content of conscious report has shifted dramatically in re- 
lation to the interpersonal purposes involved. If, however, we had 
drawn hard and fast diagnostic conclusions after the first interview, a 
most incomplete picture would have developed, A working rule thus 
develops. The data of conscious report must be studied in the light of 
the three standard sources of variation: time, the interpersonal context, 
and variation among the levels of personality. The last, which defines 
structural variation, involves the relationship between the levels of 
self-behavior. In the illustration we have just considered, we focus on 
the interpersonal context in which the parents were described. The 
patient's Level I purposes shifted from defensive disagreement to co- 
operation, as the social environment, in the form of the therapist, 
shifted its directive pressure. 

It follows that one control over the variability in conscious descrip- 
tion is obtained by indicating the organism — world matrix from which 
the data come. Since patients give us their views of self-world in the 
context in which we wish to employ the data — i.e., in the clinic — the 
functional criterion enters again as a second useful control over varia- 
bility. By pointing our predictions to future behavior in the clinic, 
we keep constant, or at least more constant, the situational factor. The 
advantage of limiting our predictions to the functional nexus has al- 
ready been considered in the preceding chapter. They are equally 
applicable to the problem of Level II variation, 

A third partial solution to the issue of variation involves technical 
procedures in the collection of data. There is a wide variety of meth- 
ods for obtaining Level II material in the clinical situation — interview, 
check list, autobiography, etc. They range from the personal revela- 
tion at the most intensive moments of psychotherapy to mechanical se- 



lection of "yes" or "no" items on a questionnaire. For a complete eval- 
uation of personality, we optimally obtain as many different types of 
self-report as possible — as the subject varies in response to the most 
free through the most controlled stimuli, from the most confiding to 
the most defensive motivations. 

In this w2Ly we tap not just the patient's self-description as revealed 
by one Level II measure — but rather a range of Level II behaviors. If 
the same self-description emerges from all the measures, then we can 
be fairly certain that we have a durable estimate of Level II. If it 
varies among the different sublevel measures, then we have an esti- 
mate of the changeability of the self-description and the way it varies. 

Figure 14 provides a hypothetical Level II variation in depth of con- 
scious reports in different cultural contexts. The problem of varia- 
bility, which we have raised here, is discussed in detail in a later sec- 
tion of this book. 

Various Level II Measures 

Level III 

Level II-C Level II-C Level Il-Ti Level II-Ti 

Check list 

of self 
during job 


Check list 

of self 
when tested 
in clinic 


of self 
at begin- 
ning of 


of self 

after one 

year of 




Figure 14. Hypothetical Variation in Depth of Level II Measurements Due to 
Change in Cultural Context. 

Incidence of Level ll-C Behavior in Various Cultural Samples 

Summaries of research findings involving Level II behavior are lo- 
cated in the pertinent clinical chapters to follow. To familiarize the 
reader with some of the general meaning of conscious descriptions of 
self and others the percentage of Level II types occurring in several 
symptomatic samples will now be presented in Table 3. 


s ss§ 






































































l/^ NO -^ rr, ^ 

O^ O OO O Tj- O r-j 

Ov vo ■^ — 

« 00 o ^ -^ 

r~~ — vo o -^ o 

"^ " .ii E 

^^ c '5 ■« ■" 

!2 « w c£ S t** 

o 2 -H Oh -a S 5 g 

'S « ji 'w p^ -n „ ^ 

-7^ n X (1 H -s 



It will be noted that in their self-diagnosis individual psychotherapy 
patients emphasize passivity (HI), group therapy patients distrust 
(FG), ulcer patients aggressiveness (DE), hypertensive and obese pa- 
tients hypernormal strength (AP and NO). 

Tests of significance among these samples and a detailed discussion 
of the implications of these findings will be presented in Chapter 24. 


1. Cutler, R. The relationship between the therapist's personality and certain aspects 
of psychotherapy. Unpubhshed doctor's dissertation, University of Michigan, 

2. Sarvis, Mary. Personal communication. 

The Level of Private Perception. 
The Interpersonal Symbol 

The third level of personality — Level III — comprises the expressions 
that an individual makes, not directly about his real self in his real 
world, but indirectly about an imagined self in his preconscious or 
symbolic world. The interpersonal motives and actions attributed to 
the figures who people his fantasies, his creative expressions, his wishes, 
his dreams define the subject matter for this level of personality. They 
are called preconscious symbolic expressions because they stand for 
or symbolize aspects of the subject (and his world) which are not di- 
rectly denoted. The subject selects and employs themes. But he at- 
tributes them not to himself or to his real world, but in an imaginary 
context. They do have a relationship — although indirect and often un- 
witting — to his conscious and communicative behavior. They have an 
expressive function, not direct, but symbolic. 

The use of the term preconscious to describe Level III expressions 
is a debatable procedure. The preconscious as defined operationally 
in this volume cannot be equated with the term as used by psycho- 
analysts. Kris (6, p. 542) has cited two quotations from Freud which 
define the preconscious. "In defining the quality of the preconscious, 
Freud follows Breuer: preconscious is what is 'capable of becoming 
conscious,' and he adds, 'capable of becoming conscious easily and 
under conditions which frequently arise.' " It might seem, at first 
glance, that the symbolic and projective responses which define Level 
III might meet Freud's definition. The empirical situation is, un- 
happily, not that simple. In actuality subjects do not always express 
in response to projective stimuli their private or fantasy thoughts. 
Many defensive, suppressive, rigid patients repeat in their response to 
projective stimuli the same themes they report in their conscious de- 
scriptions. What we get at Level III is, therefore, not preconscious ma- 



terial, but those themes which the subject is wilUng to express in the 
testing situation. Level III is defined by the source of the data. Level 
III behavior should, therefore, be accurately labeled as the "response 
to projective stimuU." This may not be indirect or symbolic or pre- 
conscious. Since there is no single term in the English language for 
denoting "that which the subject chooses to express in reaction to pro- 
jective stimuli," I have hesitantly employed the familiar terms "sym- 
bolic, imaginative, indirect, fantasy, projective and preconscious" as 
synonyms for Level III behavior. To remind the reader that the 
psychoanalytic concept is not denoted, the word preconscious will be 
consistently in quotes. 

Whenever the subject shifts the content of expression from the 
actuality — believed, perceived, described — to the imagined, fantasied, 
then he is communicating in the symbolic mode. In practice this dis- 
tinction is quite simple and straightforward. The content of dreams, 
fantasies, creative expressions, wishes, projective tests, automatically 
becomes Level III data. 

The Paradox of Symbolic Life 

The phenomenon of symbolization is one of the most puzzling as- 
pects of human behavior. In the first place it seems to be a universal 
phenomenon. The dream, that obscure, enigmatic ripple across the 
surface of rational life, is shared by the most literate and the most 
primitive mind. Rituals, legends, myths, fantasies are woven into the 
histories of all people and all cultures. 

A second paradoxical quality of symbols is their function. Why do 
all men channel so much energy into symbolization? Man's response 
to the physical elements is fairly well rationalized. We can explain 
the cognitive aspects of behavior — the communicative functions of 
sign, gesture, words in the pattern of social survival has been exten- 
sively studied. While the theories disagree in detail, the general 
purpose of representative signs, whether cries of alarm or notational 
ciphers, is an exphcable area of knowledge. Genetic and evolutionary 
theories have had considerable success in explaining the survival value 
of these communications. The discursive language of factual descrip- 
tion, which we have discussed in the last chapter, is the most elaborated 
and practical aspect of human intellectual life. The essence of this 
form of expression is that it refers to events and things in the objective 
world. This is called the representational function. 

The symbolic function is, however, quite a different phenomenon. 
It does not refer to the world as it is seen by others. It does not neces- 
sarily rely on the lawful principles that regulate the events of reality. 
Miraculous, magical processes can transpire in fantasies. The limits of 


space and time do not apply in dreams. The unreal quality of symbolic 
productions, which for some philosophers is its outstanding character- 
istic, certainly assigns them a different function in the economy of 
human life. The imaginative mode is of little direct use in dealing im- 
mediately with the practical aspects of life. Thus, as Langer points 
out, creative-autistic expressions cannot be explained in terms of their 
survival value in dealing with the real world. In many cases man's 
myths and fantastic autisms have confused and hampered his adjust- 
ment to the environment around him (7). 

The most persuasive solution to this paradox is that symbolic ex- 
pression is not a response by which man deals with the challenging 
stimuli of the external environment; it is a response to internal am- 
biguity and tension. 

Reversal Theory of Symbols 

In Chapter 7 when we discussed the level of public communication, 
considerable emphasis was placed on the reflex tendency to select cer- 
tain interpersonal responses and to avoid others. The phenomenon of 
reciprocal interpersonal relations formalized this automatic process by 
which we pull certain reactions from others and, in turn, respond with 
a limited set of behaviors. The stable continuity that thus develops 
results in an imbalance. Certain interpersonal techniques for minimiz- 
ing anxiety are automatically employed. Others which cause anxiety 
are less favored. 

In Jungian terms, certain interpersonal functions are overdeveloped; 
others are neglected. One side of the circle is predominant, the other 
inhibited. An imbalance at one level of personality can have many 
possible relationships to the rest of the character structure. The sim- 
plest and most classic case is the reversal concept repression of the op- 
posite. Here symbols are held to express the exact opposite of overt or 
conscious behavior. This is an appealing solution of the conscious- 
unconscious problem and has by far the most common sense appeal. 
According to this version, something like a conservation of energy 
process is at work. The themes which are inhibited and denied from 
overt manifestation are held to be expressed in symbolic life. Almost 
every theorist who has written on the psychology of symbolism has 
leaned on this notion. Much anecdotal evidence supports it. The 
private life of Walter Mitty is shot through with the acclaim, success, 
and mastery which he does not express in his prosaic life. 

Objections to the Reversal Theory of Symbols 

The general popularity of the reversal theory of symbolism has not 
been diminished by the two demurrers that can be raised against it. In 


the first place, there are many cases in which it just does not work — 
the nightmare dream, to take the extreme example, in which the pain 
of reality is repeated again and again in exaggerated form. A second 
restraint on the easy acceptance of this theory is that it has never had 
objective assessment. The repression of the opposite theory has never 
been put to the test because such testing requires a systematic method 
for measuring behavior at the conscious and overt levels, for measuring 
behavior at the symbolic level in terms of the same variable continuum, 
and for relating the different levels. 

The Kaiser Foundation Psychology Research Group has attempted 
to test this reversal hypothesis in a series of correlational studies. This 
research suggests that the tendency for symbols to express the opposite 
of conscious or public behavior is not universal or inevitable. It holds 
for about half of our cases, some of the time. Some persons do tend 
to employ symbols which are the opposite of their conscious and pub- 
lic imbalances, but others tend to report monotonously in their sym- 
bols the same themes which characterize their behavior at other levels. 
Thus, the reversal or equilibrium theory of fantasy is not a general 
finding. It varies from person to person. This variability, the tendency 
to use symbols which are the same or different from consciousness, is 
a measurable, stable, psychological variable. In Chapter 13 it will be 
defined and validated as a separate and vital dimension of personality 
in its own right. 

Symbols Are Important in Relationship to Other Levels 

The fallacy behind oversimplified hypotheses such as the reversal of 
sy^nbols theory is that they focus on a single level of personality — at 
best, two levels — and attempt to generalize laws. The results are bound 
to be disappointing and misleading. 

We approach the indirect imaginative productions of the human 
being not expecting them to serve any single function. Any level takes 
on its full meaning only in relationship to all the other levels, that is, 
to the total personality organization. This is the organismic assump- 
tion. Every level or area of personality is in dynamic equilibrium with 
all the other levels and the total intricate system of balance and 
counterbalance makes up the fabulous complexity we call personality. 
To prevent this organismic assumption from becoming a truism, the 
syntactical procedures determining the exact number of the interlevel 
relationships must be made explicit, then the connection between 
Level III and the other levels of personality can be defined, measured, 
validated, and understood. The permutations and combinations of 
levels according to the present notational system will be presented 
in Chapter 13. 


Some Misconceptions About Projective Tests 

These findings carry along in their wake some implications for the 
projective testing movement. This branch of clinical psychology 
concerns itself with responses to stimulus items which are unstructured 
or semistructured. The subject is given vague or incomplete test cards 
and asked to give his interpretation of them. In the Rorschach test he 
tells what he "sees" in vague inkblot shapes; in the Thematic Apper- 
ception Test (TAT) he tells stories which he believes fit and com- 
plete the actions portrayed in magazine-type illustrations. The essence 
of the technique is that the stimuli are ambiguous to some degree and 
the patient "projects" his own imaginative perceptions. He attributed 
his own fantasy themes. The theory claims that through his symbols 
the subject shall be known. The themes elicited are believed to re- 
flect a "deeper" and more valid picture of his personality than those 
of conscious report. 

The field of projective testing is a theoretical shambles. In the first 
place, interpersonal, Freudian, Jungian, and stimulus-bound variables 
are jumbled together. The diagnosis is often made in a rag-tail man- 
ner, stressing whatever variables happen to drift into focus in the 
patient's responses or in the clinician's observations. There is rarely 
any attempt to separate levels. For these reasons, most of the objec- 
tive assessments of projective test practices have come up with nega- 
tive findings. Almost every time that independent researchers have 
tested the hypotheses and predictions involved in projective testing to 
see if they really work, the answer is "no." It can be flatly said that 
the field of projective testing, whatever its popularity, is an unvalidated 
or unsatisfactorily validated enterprise. 

The great potential value of this approach to the symbolic has born 
little fruit because the systematic conceptualization of levels and 
variables of personality has not been employed. Consider, for example, 
the patient who produces fantasy materials which are saturated with 
themes of bitterness and murderous anger. What can we say about the 
person on the basis of these data? Not very much. We know that his 
symbols are hostile, but without knowledge of the other levels, our 
predictions are very limited. If this subject consciously describes him- 
self as loving-agreeable, the symbolic rage takes on one significance. 
If the bitterness assigned to his fantasy heroes is also attributed to him- 
self at Level II, quite a different interpretation results. 

Formal, Noninterpersonal Aspects of Projective Tests 

In addition to the content, another aspect of projective tests refers 
to the so-called formal qualities. Here the clinician studies not the 
themes but the expressive and stylistic factors of the subject's response. 



The impulsivity, constriction, obsessive deliberation, flexibility, care- 
lessness of the performance are observed and measured. They are then 
indicated in the diagnostic report as characteristic traits. 

These behaviors comprise a valid and important aspect of personal- 
ity. They are noninterpersonal, i.e., they refer to symptomatic, stylis- 
tic mood factors. They shift us into a dimension of personality which 
is distinct from (although lawfully related to) interpersonal behavior. 
Two critical comments appear to be appropriate in considering these 
noninterpersonal variables. They are an important part of diagnostic 
procedure. However, they take on increased meaning to the extent 
that they are systematically related to interpersonal variables. No test 
report is complete which summarizes the noninterpersonal style and 
mood aspects of behavior and fails to include interpersonal prediction. 
Noninterpersonal variables like any other personality measurements 
have meaning only in relationship to the total multilevel pattern of 
purposive behavior. The statement: "The patient acts depressed, im- 
mobilized, and constricted on the Rorschach" is a good diagnostic be- 
ginning, but it takes on considerably more meaning when we fit it into 
the broader purposive context, "His immobilized sadness is accom- 
panied by self-descriptions of weakness and helplessness (Level II) 
and by dependent pressure directed toward the clinician (Level I), 
etc., etc." 

The symptomatic, diagnostic conditions which are not directly 
interpersonal have been traditionally the central concern of descrip- 
tive or medically oriented psychiatry. The Kraepelinian clinician is 
especially interested in the peripheral area of mood (depressed, agi- 
tated, manic) or style of expression (bizarre, obsessive, disorganized, 
impulsive) . The more physiological-neurological the psychiatrist's ap- 
proach, the more you may be sure that he will avoid interpersonal 
terminology and depend on the peripheral-symptomatic. In shifting 
the emphasis to the social dimensions of personality, we by no means 
neglect the noninterpersonal. We make the hypothesis that the expres- 
sive and mood variables of personality have a basic, although indirect 
meaning. As we shall see, they are related significantly to interpersonal 
factors at different levels of personality. The symptomatic aspects 
of psychiatry thus take on an interpersonal meaning. They are related 
to interpersonal purposes. They predict interpersonal behavior. 

A second comment can be made in regard to the noninterpersonal 
variables tapped by some projective tests. It should be, but rarely is, 
kept clear that these reactions have little or nothing to do with the 
symbolic mode. They comprise an entirely separate dimension of be- 
havior. Certain expressive noninterpersonal factors are related to 
specific interpersonal themes (e.g., energeticness is related to assertive- 


ness), but two discrete dimensions of measurement are involved. Now 
the content of imaginative expressions — the themes and purposive 
motifs — comprise the language of symbolism which we study as Level 
III data. The manner in which the subject deals with the symbolic 
stimulus materials — lethargically, constrictedly, unhappily — define an- 
other level and dimension of personality. The subject can produce a 
fantasy story in which the themes involve dashing, daring, careless 
impulsivity; but the way in which he narrates his story can be de- 
liberate, plodding, and painstaking. The theme of the story can be 
quite different from the way in which it is told. In this case, the inter- 
personal content of imaginative expressions is assigned to Level III 
while ratings of the manner and style of expression are assigned to 
Level I in the noninterpersonal dimension, since they are actions ob- 
served and judged by others. 

Confusion and vagueness about levels, failure to define them, and 
neglect of logical systematization has led to this strange situation: 
many projective tests are employed to tap and study not the symbolic 
mode, but the motor, perceptual response of the subject. The exceed- 
ingly popular Rorschach test stands out as the classic example in this 
regard. The standard text on Rorschach analysis devotes over seventy 
per cent of its interpretative attention to variables that have nothing to 
do with the symbolic mode. The subject's perceptual, executive, or- 
ganizational techniques, the freedom or constriction of his "affect," 
the accuracy or deviation of his perceptions and similar topics carry 
the interpretative burden. The nature and meaning of the symbolic 
language — the content of the responses — has generally been the step- 
child of Rorschach-type theory. 

The importance of the perceptual and motor executive aspects of 
behavior should, emphatically, not be minimized. To measure these 
Level I noninterpersonal variables many straightforward testing tech- 
niques suggest themselves. Experimental psychology and aptitude 
testing procedures give any number of techniques for assessing im- 
pulsivity, organizational synthesizing abilities, perceptual and intel- 
lectual functioning. All of these techniques are free from the ex- 
' haustive stimulus-bound complexity of the Rorschach. They are also 
free from the crystal ball, medicine-man aura of the ink blot pro- 
cedure, which often lends a mysterious and untherapeutic tone to the 
clinical contact. 

To measure symbolic behavior it is necessary to focus on the the- 
matic aspect of the imaginative production. The diminishing popu- 
larity of the Rorschach and the increasing trend toward content 
analysis in projective tests are, from this standpoint, healthy develop- 


Formal Versus Thematic Interpretation of Symbols 

This distinction between form of expression and content of expres- 
sion has been made by most psychologists who have concerned them- 
selves with thematic tests. (II) The syntax of levels sharpens this di- 
vision and takes the important step of assigning the two behaviors to 
two different dimensions and levels of personality. The logical classi- 
fication which results has further implications in the broader field of 
artistic interpretation. Whenever psychologists venture to apply their 
theories to the aesthetic and creative realm they should, and usually 
do, make it clear that their analyses refer not to the form but to the 
content of the artistic production. Psychologists have made many 
brilliant expositions of the thematic meaning of creative expression, 
but they have properly said little as to the artistic or formal merit. 
They attempt to understand ivhat the artist is communicating and not 
hoiv skillfully he is expressing it. 

The logic of levels makes this distinction quite clear. Our judg- 
ments of the form, the style, the manner of behavior is a Level I op- 
eration. We are rating noninterpersonal behavior. Our judgments as 
to the meaning and thematic message being communicated is a Level 
III operation. We are rating symbolic expression. 

The Function of Symbols 

In the last few pages we have been circling around the general ques- 
tion of the meaning of symbohc activity. Symbols are not necessarily 
the reverse of the coin of consciousness as the theory of opposites 
would lead us to believe. Nor are they always the behaviors tapped 
by the so-called projective tests. We shall now consider some answers 
to the questions: "What is the function, meaning, and purpose of 
symbols, and what is their practical clinical use?" 

Symbols Are a Private "Preconscious" Language. First, it can 
be said that imaginative expressions are a form of communication, an 
indirect form. The individual does not tell us directly about himself; 
he describes a fantasy or unreal set of events. Symbols are the vocab- 
ulary of a private language. When the subject talks directly about 
himself (in Level II) he is describing himself to another person. He 
is telling the psychiatrist, the tester, or the other patients in his group 
about his perceptions of self and world. When he talks in the sym- 
bolic language of dream or fantasy he is not telling others about him- 
self or his real world. He may be, in a sense, talking to himself. Sym- 
bolic language is inexplicable and mysterious if we try to interpret it 
as though the person were talking directly and openly about his con- 


scious perceptions. When interpreted as private language, we see that 
it can be understood only in the context of its personal meaning to the 

It is very well known that all individuals have a set of private per- 
ceptions, private opinions, and private reactions which often contrast 
with the statements of conscious report. This has classically been the 
despair of philosophers who have had to concede that we can never 
know exactly what goes on in the mind of another human being. It has 
made the topic of "consciousness" the source of unending speculative 
frustration. The first step in approaching this riddle is to accept the 
inevitable limitations and indeterminacy involved. The second step is 
to develop the best means for getting as close as possible to the "pre- 
conscious" or private world of our fellow men. At the present time, 
this can be best accomplished through the language of symbolism. ' 

This is by no means a simple or unambiguous procedure. In many 
cases the subject is made quite anxious if he attempts to translate his 
own private expressions into the language of direct conscious descrip- 
tion. In many cases he is made even more anxious at the prospect of 
others approaching his idiom. The expression and interpretation of 
symbols is loaded with complicating qualifications. Their meaning al- 
ways depends on the dynamics of the total personality and of the 
context in which the symbols are expressed. Thus the level of per- 
sonality which includes imaginative indirect communications offers, 
on the one hand, the most promising avenue to the private world of 
the subject, and involves, on the other hand, the most ambiguity and 
interpretive uncertainty. 

Symbols Reduce Anxiety. Next we must consider the function 
of symbolic behavior. Why do human beings develop private lan- 
guages? The first answer to this question seems to follow quite logi- 
cally; they develop indirect behaviors to avoid the anxiety of the direct. 
They express certain themes privately to avoid the anxiety that pub- 
lic expression would entail. 

We have seen in earlier chapters that all individuals develop auto- 
matic interpersonal response preferences. They use some favored re- 
actions and avoid others which would involve greater anxiety. By 
means of the language of symbolism it is possible to express inter- 
personal themes that are inhibited from direct expression. We are re- 
turning here to the old principle of expression of the opposite which, 
we have learned from the data, works only part of the time. We must 
complete the explanation by adding the other end of the continuum: 
by means of the language of symbolism it is also possible to repeat and 
thus strengthen the same themes that are manifested in direct expres- 


sion and to avoid further the themes that are inhibited from direct ex- 
pression. The purpose of symbolic behavior is to reduce anxiety. 
For some individuals this is accomplished by employing fantasy as a 
safety valve, an opportunity to "blow off" the interpersonal steam that 
has built up through inhibitions and repressions. For others, even in- 
direct, imaginative expression of the inhibited themes is anxiety-laden. 
SymboHc behavior in these cases becomes a way of strengthening the 
avoidance maneuvers. 

Symbolic Mode Indicates the Source and Amount of Anxiety. 
Thus we see that there is no simple, one-way explanation for the mean- 
ing of symbolic language that works for all cases. We lose the com- 
forting simplicity of a generalized rule. But we gain, instead, a new 
illuminating hypothesis: symbolic language can serve as an index as to 
the amount and source of the subject's anxiety. The patient who 
rigidly limits his direct interpersonal activity at Level I to a few nar- 
rowed responses and avoids all others can go on to develop any num- 
ber of symbolic resolutions. If the rigid limitation continues in his 
imaginative productions then we can assume that the anxiety which 
cripples and inhibits the absent interpersonal themes is so intense that 
he cannot express them even indirectly in the private language of 
symbols.'^ If, on the contrary, the themes which are avoided in con- 
scious report or pubhc communication appear at the level of private 
conversation, we can make the hypothesis that the anxiety is less crip- 
pling and that increased flexibility, mobility, and potential for change 
exist. And we have, further, a clue as to the direction of the antici- 
pated change, as indicated by the new themes that appear in the priv- 
ate language of symbols. These hypotheses bear up under the ob- 
jective test (see Appendix 3). 

Symbols Can Express Underlying Feelings of Uniqueness and 
Self-Consolation. We interpret Level III productions in the light of 
the total personality. It is obvious that symbols can present the same 
thematic picture as the other levels, they can be "more so," or they can 
be different. If the latter is true, then the individual has expressed in 
fantasy the themes he has inhibited in public communications. He may 
say publicly, "I am meek and weak and suffering." Now he may add 
the private comment, "but I am also concerned with the theme of re- 
taliation, or power, or prestige." To present this illustration in other 
words, the individual is saying: "I tell you openly that I am submissive, 

^ The cultural situation in which the symbols are expressed is a crucial factor. 
Powelson and Bendix (9) have described the effect that a punitive, custodial environ- 
ment can have on patients' behavior. The cultural context must be added as a qualify- 
ing variable to all the generalizations made in this chapter. 


but I wish, or hope, or symbolically perceive myself to be strong and 
powerful." Self-esteem is increased and anxiety diminished by the 
secret fantasy of fearful power. 

This kind of self-punitive masochist in fantasy asserts himself and 
retaliates against his tormenters. This is a familiar tune. It illustrates 
the notion of reversal. We have designated this as an interlevel con- 
flict — masochism at Level II opposed to sadistic assertion at Level III. 
But how about the masochist whose fantasy productions are saturated 
with even more self-defeat? How is this explained in terms of warded 
off anxiety and the theory of stabilization? These persons — and there 
are many of them — are convinced at all levels of expression that suf- 
fering and self-abasement is the safest, least threatening method of ad- 
justment. They are, we assume, less anxious when they express maso- 
chistic themes. They appear to get some consolation, excuse, and 
poignant merit from unhappy fantasies. 

These patients seem to be saying something like this: "I am overtly 
meek, weak and suffering . . . and covertly I do not perceive my- 
self as anything different." Often the pessimistic corollary is: "I dare 
not change or I do not wish to change." Patients may enhance their 
esteem and feelings of uniqueness by means of their private symbols: 
"No one is as uniquely unloved, helpless, and martyred as I am." 

The conceptualization of masochistic behavior is traditionally the 
proving ground on which personality theories meet their most taxing 
tests. It is the point where the logical assumptions based on survival 
value begin to buckle and where new concepts, such as death instincts, 
are classically dragged into action. We have attempted in the last few 
pages to employ a motivating principle — avoidance of anxiety and 
preservation of self-esteem — to explain symbolic activity in the same 
terms as public behavior. 

Time-Binding Nature of Symbols. With these remarks as pre- 
liminary it is now possible to present the essential point of this chapter. 
Symbolic, indirect or ''''pre conscious" activities are necessary for the 
human being because he is a time-binding individual. 

Unlike most other mammals, the human being continually faces and 
deals with conflictful situations in which anxiety threatens in at least 
two directions. The interpersonal world he has created pushes him 
toward one set (and often an imbalanced set) of anxiety reducing be- 
haviors. The pressures toward flexibility, both cultural and personal, 
may push him toward another source of self-esteem. The individual's 
overt behavior does not express the impulse or desire which he feels. 

This point has been well made by Murray and Kluckhohn (5 p. 18). 
They point out that: ". . . the personality is almost continuously in- 


volved in deciding between alternative or conflicting tendencies or 
elements." Personalities deal with these conflicts by constructing 
"schedules which permit the execution of as many connotations as pos- 
sible, one after the other." They go on to say, "Most men are forced 
by circumstances to make decisions which commit them to schedules 
arranged by others (e.g., the daily routine of a job); and so a large 
portion of the temporal order of their days is not of their own shap- 
ing. Also, every culture prescribes schedules, general and special, 
which define the proper time, place or order of certain actions, and, 
therefore, schedule-making is a sphere in which the individual is likely 
to come into conflict with his society." 

Postponement of impulse is thus an inevitable characteristic of hu- 
man behavior. The individual is continually inhibiting some actions 
in favor of others, generally moving in the direction of the lesser anx- 

This postponement phenomenon is called the time-binding aspect 
of human behavior. The function of "preconscious" or "unconscious" 
activities might thus be explained as time-binding. The basic discovery 
of Freud that unexpressed impulses do not disappear but remain as 
active, although indirect, elements in the personality can be considered 
as a temporal rather than a structural phenomenon. The unexpressed 
motives relate to the past and the future. 

From the functional viewpoint, the essence of private or "precon- 
scious" factors is that they are potentials for later overt or at least 
conscious expression. If they did not have this potentiality then they 
have little meaning. This time-binding theory of the "preconscious" 
is important and useful in the interpretation of symbolic behavior. 

Whenever we obtain a symbolic, "preconscious" theme from a sub- 
ject, it suggests that this theme is a potential for future action. The 
time-binding theory of the "preconscious" places the symbol pro- 
duced in the present on a temporal dimension pointing (we assume) 
to earlier frustration and functionally more important to a later ex- 
pression of the theme. 

Symbols Predict Future Behavior. The functional value of sym- 
bolic behavior to the clinician can now be stated. The data of Level 
III are predictions of the future. We have proposed the hypothesis 
that the patient's symbols tell us, in the case of conflict between con- 
scious and "preconscious" themes: "Here is another side of my inter- 
personal picture." In the case where the fantasy themes are not dis- 
crepant from overt behavior he tells us: "My rigid pattern does not 
change even in symbols." The "preconscious" themes tell us how 
likely the person is to change his behavior and in what direction he is 


likely to change. The usefulness of this information in clinical prac- 
tice is obvious. 

If this theory is correct, then symbols should indicate the amount of 
anxiety that operates in any given personality structure, and they 
should give us an estimation of the amount of change to be expected 
and the type of change to be expected. 

If this theory is correct, a prognostic instrument of considerable 
importance becomes available to clinicians. And conversely, if the 
predictions do hold up when applied to clinical practice, a major theo- 
retical step will have been taken in explaining that area of personality 
which has always been so resistant to explanation. In helping the 
clinician predict, we shall have validated the theory of symbols. 

In order to test this hypothesis, several methodological problems 
had to be met. It was necessary to convert the loose, diverse language 
of symbols into scientific categories. The interpersonal variables thus 
defined must be capable of reliable measurement. They must be di- 
rectly and systematically related to the other levels of personality so 
that interlevel conflicts and discrepancies can be measured. 

The data for Level III, it will be remembered, are defined auto- 
matically by the source from which they come. A dream, a fantasy, or 
any projective and imaginative expression reported by a subject is 
assigned to the symbolic mode. The "preconscious" level, like the 
more overt levels, is divided into two areas: self and other, or symbolic 
hero and symbolic nvorld. This division produces two distinctly dif- 
ferent types of Level III material which have unique applications and 
lawful relationships to other levels of personality. 

Six Methods for Measuring Level III Behavior 

There are six methods which have been employed by the Kaiser 
Foundation project for measuring Level III behavior. Whenever the 
discussion centers on the generic level of "preconscious" expression we 
used the code Level III. Whenever we refer to specific measurements 
of "preconscious" behavior it is necessary to indicate the specific 
source of the data by adding the appropriate code letter. This is ac- 
complished as follows: 

When trained personnel rate the interpersonal content of responses 
from the Iflund projective test (4) the scores are labeled Level lU-i. 
Scores for the Blacky projective test (1) are coded Level lll-B. 

When the interpersonal themes for dreams are rated by trained per- 
sonnel the scores are indicated as Level lll-D. Themes from waking 
fantasies are coded Level lll-F. 

MMPI indices which are being developed to predict preconscious 
behavior are coded Level lll-M. 


When trained personnel rate the interpersonal themes from the 
Thematic Aperception Test (TAT) the scores are labeled Level lll-T. 
The research studies reported in this book employ the TAT as the 
standard instrument for Level III. 

When trained personnel rate the interpersonal themes from the 
Interpersonal Fantasy Test the scores are coded Level III-IFT.^ 

The hypothesis that "preconscious" symbols predict future be- 
havior to be expected at overt or conscious levels has been tested by 
means of several research studies. The design and detailed results are 
presented in Appendix 3. These findings do confirm this hypothesis. 
"Preconscious" behavior does predict the kind and the amount of 
change to be expected in future overt behavior. If the TAT, for ex- 
ample, is more hostile than the self-diagnosis at the first testing, then 
the self-diagnosis can be expected to change over time in the direction 
of greater hostility. 

The methodology for measuring symbolic behavior which is used 
in clinical diagnosis and in these validating research studies will now be 

The Measurement of Interpersonal Symbols 

The raw data of Level III comprise the verbal language of the 
dream texts, fantasy stories, projective test protocols. The task here 
is the same as at any other level of personality — to convert the raw 
protocol language into objective categories. One of the basic prin- 
ciples of the interpersonal system of personality is that the same matrix 
of variables should be used at every level of behavior. This facilitates 
direct comparison between levels. 

This principle settles in advance the issue of what variables are to be 
used in measuring symbolic behavior. The sixteen-variable circular 
continuum by which we rate public communications and conscious 
descriptions is also employed to analyze the imaginative data. 

Classification of Fantasy Materials into ''Hero'' and ''Others'' 

We rate the heroes and protagonists of the fantasy world just the 
way we rate the activity of the subject's real self and real ivorld at 
Levels II and III. If the dream hero is fearful, he is scored H; if he is 
murderously enraged, we score E, etc. The same procedure of differ- 
entiating between the self and the other is preserved. The hero of the 

^ The Interpersonal Fantasy Test is a projective instrument developed to tap the 
subject's fantasy descriptions of the permutations and combinations of the most stand- 
ard and crucial interpersonal relationships. It is tailored to the interpersonal system 
just as the Interpersonal Check List is designed to tap the sixteen variable continuum. 
The Interpersonal Fantasy Test is published by the Psychological Consultation Service, 
Berkeley, California. 


dream becomes the symbolic self. The people he interacts with de- 
note the interpersonal other. The identities of these figures of the 
interpersonal world are specified much as they are at Levels I and II. 
Formal rules and conventions for determining which character in a 
dream or fantasy story is the hero and which are the "others" have 
been developed. (See Appendix 3.) 

This division into "preconscious" hero and "preconscious" other 
is an important one. It defines two separate sublevels of the "pre- 
conscious." Studies taken from several widely differing samples con- 
sistently show that the "preconscious" other scores are significantly 
different from the "preconscious" self or hero scores. 

The establishment of an operationally defined hierarchy of "pre- 
conscious" layers has functional value. It helps us understand the 
amount of anxiety connected to any interpersonal behavior. If an 
emotion — let us say, rebelliousness — appears at Level II, we can as- 
sume that the subject is not made so anxious by the emotion that he 
must avoid it consciously. He can tolerate this interpersonal behavior 
in conscious report. If he completely avoids or denies it at Levels 
I and II, one assumes its direct expression makes him anxious. 

If, in this latter case, the rebelliousness appears at Level III hero, 
this indicates that he can tolerate the emotion at the "top layer of 
fantasy." This indicates that the anxiety connected with unconven- 
tional behavior is not too massive or crippling. 

Let us consider the case where rebelhousness is completely avoided 
at Levels I, II, and Level III hero. This suggests that the three most 
overt levels are organized against the expression of the emotion. 
Considerable anxiety must be connected with its expression. 

At this point the distinction between Level III hero and other be- 
comes functionally useful. Since there is evidence that Level III other 
is deeper than Level III hero, we look to the themes attributed to the 
fantasy world. If they, too, are marked by an avoidance of rebellious- 
ness, we have an added cue as to the amount of anxiety attached. If 
the themes warded off at the three more overt layers finally appear in 
the fantasy world, then they become potentially available for future 
integration into awareness. 

This brings us, of course, to a classic item of clinical folklore — it is 
easier for the patient to master and integrate feelings which are pro- 
jected onto others than if they do not appear at all. Pulling back 
projections and accepting them into the self-structure is a basic thera- 
peutic procedure. The differentiation of layers of "preconscious" 
behavior is a useful device with considerable cHnical application. 
Preconscious themes attributed to fantasy others are thus considered 
part of the subject's personality, usually related meaningfully to the 
conscious or preconscious self. 


Classification of Fantasy Images 

There is a further differentiation of "preconscious" behavior which 
has certain theoretical and chnical implications. The fantasy stories 
are first divided into self or hero and the themes attributed to each are 
scored according to interpersonal variables (see next section). After 
the main hero-world themes are scored, then each character is classi- 
fied according to his familial status. The categories employed are 
maternal figures, paternal figures, cross-sex figures, and same-sex fig- 
ures. It is then possible to add up the scores which summarize the re- 
lationship between: 

Male child vs. Maternal figure 

Male child vs. Paternal figure 

Female child vs. Maternal figure 

Female child vs. Paternal figure 

Fantasy figures of subject's sex vs. Cross-sex figure 

Fantasy figures of subject's sex vs. Same-sex figure 

These procedures give a summary of the interpersonal behavior as- 
signed to these important role relationships at the level of fantasy. In 
a later chapter we shall consider indices which systematically link these 
interlevel processes, e.g., the similarity or difference between the con- 
scious and "preconscious" views of maternal figures. The classifica- 
tion of "preconscious" personages make possible the operational defi- 
nition of such processes as displacement, "preconscious" identification, 
and the like. 

Symbols Involve Greater Violence and Intensity 

There is a difference between the rating of Level III themes and 
overt behavior. The actions and traits expressed in the symbolic mode 
are usually much richer than those of the other levels. More intensity 
of feeling and violence of action occurs. Patients rarely describe mur- 
ders or world-shaking power motives in their conscious reports about 
their real lives. In their Level I observed interactions in group therapy, 
the patients may insult or help each other — but blood never flows. 
The acts of generosity are limited and generally expressed verbally. 

This is, of course, not so in fantasy. Intense affect, crime, rape, sui- 
cidal grief, physical and material generosity, bodily exchanges of love 
and hate often occur. The same interpersonal motives appear to 
underlie the interactions of both levels. The sixteen-variable con- 
tinuum seems adequate to categorize the behavior. It is the manner in 
which the interpersonal purpose is expressed that is different. 


For this reason additions to the rating system have been made in 
order to handle symbolic data. The list of specific ways in which the 
same interpersonal purposes can be manifested must be expanded. In 
a therapy group or discussion group the motive D can be expressed 
through sarcasm, disapproval, punitive comments, derision, verbal 
threats, etc. In fantasy the motive D can be expressed in these same 
terms, but in addition in a more intense manner. These generally in- 
volve physical or material modes, brutal punishments, actual destruc- 
tion, incarceration, execution, etc. 

The illustrative key employed in analyzing symbolic activity, 
therefore, includes all the actions measured at Levels II and I — but it 
also includes the more intense and violent activities of the fantasy 

The Variables of Level III 

Table 4 presents a hst of sample behaviors as rated at Level III. It 
must be kept in mind the list of verbs used at Level I and the list of 
attributes used at Level II also apply in the symbolic mode. We have 
not duplicated these lists of behaviors here, but have listed just the 
interpersonal events which are unique to Level III. 


Illustrative Classification of Interpersonal 
Behavior at the Symbolic or Projective Level 

A. The code A is assigned to themes of Power: Leadership, Command, Direction, 

B. The code B is assigned to themes of Narcissism: Independence, Self-Expression, 
Superior, Power Struggle. 

C. The code C is assigned to themes of Exploitation: Seduction, Rape, Rejecting, 
Depriving, Selfishness, Keeping Away From, Keeping Children to Self. 

D. The code D is assigned to themes of Punitive Hostility: Punishment, Coercion, 
Brutality, Quarreling, Threat. 

E. The code E is assigned to themes of All Forms of Pure Hostility: Disaffiliation, 
Murder, Anger, Fighting. 

F. The code F is assigned to themes of Unconventional Activity: Passive Resistance, 
Rebellion, Generic Crime versus Authority, Pure Jealousy, Drunkenness, Stealing 
Covertly, Offended, Bitterness. 

G. The code G is assigned to themes of Deprivation: Distrust, Disappointment, Re- 
jectedness. Suspicion, Bad Things Are Done to One. 

H. The code H is assigned to themes of Masochism: Grief, Suicide, Withdrawal, 
Guilt, Provoking Punishment, Self -Punishment, Fear, Anxiety, Insanity (Unspeci- 
fied), Loneliness, Running Away. 

I. The code / is assigned to themes of Weakness: Obedience, Submission, Uncon- 
sciousness, Indecision, Ambivalence, Immobilization, Illness, Passivity. 

J. The code / is assigned to themes of Conformity: Accepting Advice, Provoking 
Advice, Being Student, Docility, Followership, Positive Passivity. 

K. The code K is assigned to themes of Trust: Cling, Good Things Come to One, 
Good Luck, Being Taken Care of. Dependence, Gratitude. 

L. The code L is assigned to themes of Collaboration and Agreeability: Congeniality, 



Cooperation, "Generic Happy Ending" Caused by People Working Things Out, 
Adjustment in General. 

M. The code M is assigned to themes of All Forms of Pure Love: Affiliation, Mar- 
riage, Friendship. 

N. The code N is assigned to themes of Tenderness: Support, Kindness, Encourage- 
ment, Solace, Pity. 

O. The code O is assigned to themes of Generosity: Help, Curing Someone, Taking 
Care of Someone, Giving. 

P. The code P is assigned to themes of Success: Heroism, Popularity, Acclaim, 
Achievement, Wisdom, Teaching, Explaining. 

The cautionary statements made when we listed sample themes for 
Levels I and II must be repeated again. This table of themes is illus- 
trative. Actually the list of potential symbolic themes is exhausted 
only by the seemingly infinite variety of man's autism and creativity. 
We have included here only the most common. (A glance at Table 4 
makes apparent the great variety of behavior tapped in the symbolic 
mode. Any aspect of human experience can appear at Level III: sexual 
events, relations to authority, law, nature, occupational and political 
adjustments, the vicissitudes of childhood, mating, marriage, parent- 
hood, of growing up and growing old.) The illustrative words in 
Table 4 tap only a small fraction of the potential. We have concen- 
trated, therefore, on the events most commonly obtained in response 
to a standard projective test — the Thematic Apperception Test. 

In actual practice the rating is assigned not in a routine, check-off 
fashion, but by a judgmental application of the circular concept. 
That is, the list of themes presented in this illustrative figure is not ap- 
plied automatically. Murder can be scored as F, E, or D depending on 
the power element involved in the hostility. Success is B if it involves 
superiority or proving someone else inferior; it can be F if it involves 
the notion of respect or admiration from others. 

Here are examples of the interpersonal system applied to three 
varieties of Level III data: a dream of Sigmund Freud (Level III-D), 
a fantasy (Level III-F), and a projective test story (Level III-T), 

The Scoring of Interpersonal Symbols 
from a Dream of Sigmund Freud 

Freud in The Interpretation of Dreams (2, pp. 195-96) presents 
a protocol which can be used to illustrate the scoring of interpersonal 
symbols. This dream concerns "preconscious" transference-counter- 
transference feelings on the part of Freud and provides an example of 
how Level III self and other scores are employed to yield different 

Freud prefaces this dream with the explanation that Irma was a 
patient whose analysis ended only "in partial success." "I expected her 



to accept a solution which did not seem acceptable to her." Later a 
friend reported to Freud that Irma "was not quite well." Freud re- 
ports he was annoyed by the possible reproach in the friend's voice 
"that same evening I wrote the clinical history of Irma's case, in order 
to give it, as though to justify myself. . . ." 

These Level II conscious descriptions of Freud would be scored 
as follows: 


Protocol Description Self 

"I expected her to accept a solution which A or B 

did not seem acceptable to her." F 

A friend reproaches Freud. D 

This "annoyed me." E 

The patient's relatives did not approve of 
the treatment. D 

Freud tries to justify himself. B 

The relationship between Freud and the patient is summarized as: 
Freud = A or B < — > Patient = F 

The relationship between Freud and the friend and relatives of the 
patient is: 

Freud = E and B < — > Friend = D 

Freud's conscious description of self in this episode locates in the 
upper left-hand quadrant of the interpersonal circle. He is strong, 
right, and righteously angry. 

The night following these events Freud had a dream. We shall 
consider the first secrion of the dream which relates to the relationship 
to the patient. 

Dream of July 23-24, 1895 



A great hall— a number of guests, whom we are re- 
ceiving—among them Irma, whom I immediately take A— directs 
aside, as though to answer her letter, and to reproach D— reproaches 
her for not yet accepting the "solution." I say to her: 
"If you still have pains, it is really only your own 
fault."— She answers: "If you only knew what pains I 
have now in the throat, stomach, and abdomen— I am 
choked by them." I am stanled, and look at her. She 
looks pale and puffy. I think that after all I must be 
overlooking some organic affection. I take her to the A— directs 
window and look into her throat. She offers some 
resistance to this, like a woman who has a set of false D—is critical 
teeth. I think, surely, she doesn't need them. 

The scoring of this dream indicates that Freud's "preconscious' 
view of his relationship with this patient is as follows: 

F— complains 
I— acts weak 




Freud's "preconscious" self = Strong, right, punitive 
Freud's "preconscious" other = Weak and rebelUous 

At this point we have data from three layers indicating Freud's 
reaction to this interpersonal situation. At Level II Freud is strong 
and self-confident. At the next deeper level of personality, i.e.. Level 
III-D hero, he is strong and righteously angry. At Level III-D other, 
he attributes to others rebellious and weak themes. This last statement 
requires comment. We no longer consider Level III other as being a 
characteristic of the "other one," but as a deeper estimate of the sub- 
ject's own feelings which have been projected onto "preconscious" 
images. The Level III other can be seen as an underlying identifica- 
tion with weakness. 

A three-layer summary of Freud's reactions to a rather difficult in- 
terpersonal situation thus results. We see that at two top levels (Level 
I is, of course, not available) Freud was maintaining a position of 
strength, self-confidence, and sternness. At the deeper layer (Level 
Ill-other) there is evidence suggesting that Freud felt resentful and 
threatened by this experience. His rebeUiousness and feelings of weak- 
ness, we assume, were accompanied by some anxiety because they ap- 
pear at Level III-D other, projected onto the image of the patient. 
This anxiety was not crippling as indicated by Freud's frank descrip- 
tion of his feelings. 

A classic example of fantasy behavior is found in the short story 
"The Secret Life of Walter Mitty" by James Thurber (10). A sum- 
mary of his overt behavior is contained in the following episode: 

Level I-R Behavior of Walter Mitty as Scored by the Interpersonal System 


F— passively resists 


F— passively resists 

Walter Mitty stopped the car in front of the 
building where his wife went to have her hair 
done. "Remember to get those overshoes while 
I'm having my hair done," she said. "I don't 
need overshoes," said Mitty. She put her mirror 
back into her bag. "We've been all through 
that," she said, getting out of the car. "You're 
not a young man any longer." He raced the 
engine a little. "Why don't you wear your 
gloves? Have you lost your gloves?" Walter 
Mitty reached into a pocket and brought out 
the gloves. He put them on, but after she had 
turned and gone into the building and he had 
driven on to a red light, he took them off again. 
"Pick it up, brother!" snapped a cop as the 
light changed, and Mitty hastily pulled on his 
gloves and lurched ahead.^ 


A— directs 

B— patronizes 

D— ridicules 
A— directs 
D— accuses 

A— directs 

I— obeys 

These interactions are profiled in Figure 15. 

3 "The Secret Life of Walter Mitty." Copyright, 1939, James Thurber. Originally 
published in The New Yorker. 



Figure 15. Diagrammatic Summary of Walter Mitry Interaction at Level I-R. Key: 
Radius of circle = 4 raw scores. 

These profiles indicate that submissioa and passive resistance 
characterize Walter iMitty's overt behavior. He provokes bossy, su- 
perior, and critical behavior from others. 

The underlying feelings of Thurber's hero are, of course, quite dif- 
ferent. There are five fantasies included in this story each of which 
portrays the hero as commanding, successful, proud, disdainful, and 
deeply respected. 

Level Ill-F of Walter Mitty as Scored by the Interpersonal System 
(Molecular Rating) 

A— commands 

"We're going through!" The Commander's 
voice was like thin ice breaking. He wore his 
full-dress uniform, with the heavily braided white 
cap pulled down rakishly over one cold gray eye. 
"We can't make it, sir. It's spoiling for a hur- 
D— stem firmness ricane, if you ask me." "I'm not asking you, 
Lieutenant Berg," said the Commander. "Throw 
on the power hghts! Rev her up to 8,500! We're 
going through!" The pounding of the cylinders 
increased : ta-pocketa-pocketa-pocketa-poc^era- 
pocketa. The Commander stared at the ice form- 
ing on the pilot window. He walked over and 
twisted a row of complicated dials. "Switch on 
No. 8 auxiUary!" he shouted. "Switch on No. 8 
auxiliary!" repeated Lieutenant Berg. "Full 
strength in No. 3 turret!" The crew, bending 
to their various tasks in the huge, hurtling eight- 
engined Navy hydroplane, looked at each other 
and grinned. "The Old Man'll get us through," 
thev said to one another. "The Old Man ain't 
afraid of HeU!" . . .=» 

B— confidence 

A— commands 
B— confidence 

A— commands 



F— complain 

I— obeys 
I— obeys 

]— admire 


The item by item scoring of each interpersonal action in this fantasy 
illustrates the molecular system for rating Level III-F behavior. 
Where several fantasies are available it is often economical to assign a 
single summary rating to the behavior of the hero and other. This is 
called molar rating. The molar scores for this episode would be: 

Mitty Others 

AB 1} 

This scoring translates into the verbal summary: "The hero is com- 
manding and self-confident; others obey and admire him." 

The Scoring of Interpersonal Symbols 

from the The?natic Apperception Test 

In the Thematic Apperception Test (8) the subject is given a series 
of cards which contain magazine-type illustrations. The task is to tell 
a story which fits the picture. He is requested to describe the action, 
the feelings of the characters, and the outcome of the plot. 

The following illustrative story was told in response to TAT Card 
2-i which pictures a farm scene, a young woman in the foreground 
holding books, an older woman in the background leaning against 
a tree, and a man in the distance plowing. A patient told the following 
story of this picture. The scored themes are italicized: 

Symbolic Symbolic 

Self (Hero) Other 

Scores Scores 

They probably got up and had breakfast— the Mother 

and Daughter. 
The daughter looks like she hates her Mother. £-3 

I guess because of the Mother's stern look. D-2 

Her body shows no emotion. I think she'll run away H-1 

—but not from the school— because she resents her F-3 

Mother who is so unkind to her and D-3 

doesn't show her any love. C-2 

My own feeling tells me she flees from the situation. H-3 

It will be noted that every interpersonal feeling or action in this 
story was given a separate score. This method of rating every inter- 
personal unit is called molecular scoring — every interpersonal detail 
gets a separate rating. It is possible to add up all the scores given to 
the hero and to the "other" in this story — or to summarize all of the 
molecular themes from all the TAT cards. 

The scores for this particular story, combined into a Level III-T 
profile, are shown in Figure 16. This graphically portrays the themes 
of anger (£), resentment (F), and retreat (H) from a hostile (D) 
and rejecting (C) parent. 




Figure 16. Diagrammatic Representation of Interpersonal Scores for an Illustrative 
TAT Story. Key: Radius of octants of circle = 4 scores. 

Another method for scoring TAT data which has proved more 
convenient is to read over the entire story and to assign an over-all 
score to the basic relationship between hero and other. This is called 
summary or molar TAT scoring. For this particular story the molar 
rating would be: 

Hero (Daughter) 

Other (Mother) 

This summary formula tells us that the fantasy hero resents and then 
withdraws from an unsympathetic mother. 

Case SumTnaries Illustrating the Relationship 
Principle of Symbolic Data 

At this point we have just about completed the description of three 
of the four levels of personality which are now employed by the 
Kaiser Foundation project. It is now possible to put the data for the 
three levels together and to work out some tentative formulations 
about some of the relationships among levels. The following case 
studies are designed to show that symbolic data are useless when 
studied in isolation and that they are most useful when studied in re- 
lationship to other levels of personality. 

Symbols of Distrust, Deprivation, and Isolation. As part of the 
initial evaluation process at the Kaiser Foundation Psychiatric Clinic 


patients are given a battery of tests which measure different levels of 
the interpersonal system. When the tests from each level are scored, 
the results are fitted together to determine the interpersonal diagnosis 
and to provide the data for clinical prediction. 

Let us consider a subject whose "preconscious" language is loaded 
with themes of deprivation, distrust, and loneliness. One such patient 
described all his heroes as failures, unloved, beaten, and exploited. 
They all ended up suicidal, isolated, frustrated. All his endings were 

With these data in mind, what predictions can we make about his 
conscious view of self or his interpersonal reflexes? Many such pa- 
tients see themselves consciously in the same way. They report their 
real life as being frustrated, unhappy, isolated. They may describe 
their interpersonal world as being reciprocally rejecting and disap- 
pointing. In Figure 17 we see such a patient. The subject's conscious 
description matches the discouragement of his symbols. The fantasied 
others are unsympathetic, much as he describes his own father. 

Many other subjects whose symbolic heroes are deprived and iso- 
lated show a markedly different pattern at Level II-C. The self de- 
scriptions, instead of being weak, isolated, and pessimistic, may empha- 
size strength and success. A clear conflict exists between the self- 
perception and the self as symbolized. 

The meaning of the fantasy productions varies considerably in these 
two cases. The passive deprivation means one thing when it dupli- 
cates the pessimism of Levels II and I. It means another thing in the 
context of overt overoptimism and expansiveness. A diagnosis of con- 
flict is defined in the latter case. If this conflicting pattern is confirmed 
by the other system measurements and by the clinical history, a con- 
siderably different prognosis and therapeutic program would result. 

Level III takes on meaning in relationship to the other levels of per- 

Conflicting Symbolic Themes of Power and Weakness. The 
two cases just considered possessed rather narrow symbolic patterns 
emphasizing the themes FGHI. This overloading of symbolic themes 
in one direction is not unusual — neither is it inevitable. Many patients 
reflect a more ambivalent picture at the symbolic level. In these cases 
the fantasy material may indicate the nature of an underlying conflict 
— but they do not tell us which side of the ambivalence is being ex- 
pressed at the overt levels. They do not tell which interpersonal re- 
flex pattern is employed. 

Let us take as example a patient whose TAT stories were concerned 
with the themes of strength and weakness. In some of his fantasies the 







Figure 17. Level II-C and Level III-T Hero and Other Profiles for Illustrative Case. 
Key: Radius of Level III-T circles = 8 scores; radius of Level II-C circles = 16 scores. 
Raw scores are employed in these illustrative diagrams. 

heroes were wise and respected figures winning the attention of ad- 
miring followers. In other stories the heroes were docile, dependent 
figures looking up to powerful, esteemed others. All the people de- 
scribed in his imagination were either leading or being led. They 
diagrammed in Figure 1 8 : 

These symbols clearly reveal a preoccupation with strength and 
weakness. They suggest that a marked ambivalence exists concerning 
power motives. With this knowledge of the subject's fantasy con- 
flict can we diagnose the other levels of personality.^ Not very well. 
We can safely guess that the overt levels of behavior will fall along 





Figure 18. Diagrammatic Representation of Symbolic Hero and "Other" Scores 
Illustrating Conflict Between Power and Weakness. Key: Radius of each octant of 
circle = 8 scores (i.e., raw TAT ratings). 

the power-submission axis {AP versus H, I, J, K). It is statistically 
unlikely that he will describe himself or present himself as hostile or 
affectionate. We can thus eliminate roughly one half of the circular 
continuum (D, E, F, G and N, O, L,M). 

What we cannot do is predict whether he will present himself as 
strong or weak. His interpersonal reflexes may reflect either aspect 
of the underlying conflict. They may express both sides. When a 
conflict between dominance and submission exists at the "precon- 
scious" level, about one half of such cases manifest themselves as strong 
people denying weakness. Such patients resist psychotherapy, com- 
plain of physical rather than emotional symptoms, deny emotional dis- 
turbance, depression, or passivity. These patients express one side of 
their ambivalence— the strong side. They "sit on" the weak aspects. 
These patients are often called counterphobic or compensatory cases. 
This means that they react against their underlying feelings of weak- 
ness by appearing very strong, indeed. Figure 19 shows how one such 
patient appeared at Levels I-S and II-C. 

But many other patients with the same TAT pattern of conflicting 
strength and weakness appear quite differently at the other levels. 
They stress the passivity side of the power-passivity axis. They claim 
to be depressed, immobilized, inferior people. They apply for psycho- 
therapy eagerly. They have strongly developed reflexes of helpless- 
ness. They willingly admit their need for treatment. These people 
are also "sitting on" one side of their underlying conflict— the strong 
side. The fantasy themes of power and esteem are consciously attrib- 





Figure 19. Diagrammatic Representation of Level I-S & II-C Self Scores Illustrating 
a Facade of Power and Responsibility. 

u ted to Others (often the therapist) . The strong-dominant side of their 
fantasy coin which does not overtly appear usually manifests itself in 
the later sequences of the relationship. Patients who present them- 
selves in this manner are commonly called by several names (usually 
depending on the specific, peripheral aspects of symptomology) . 
They include the phobics, neurasthenics, anxiety neurotics. Figure 20 
shows how a typical case might profile. 

' (Hi 



Figure 20. Diagrammatic Representation of Level I-S & II-C Self Scores Illustrating 
a Fagade of Weakness and Docility. 


Two theoretical points have been developed in this last series of il- 
lustrations — one old, the other new. The first is the familiar refrain 
— the language of symbols does not necessarily duplicate or reverse the 
other levels of personality structure. In his overt behavior the patient 
may repeat the symbolic motifs — or he may be counterbalanced away 
from them. Symbols often predict future change in overt behavior 
— but their meaning must always be assessed in terms of the total per- 
sonality structure. 

The second point concerns the notion of the conflict axis. We have 
noted that symbols can be the same or they can be different from the 
levels of reflex action and perception. They can also be mixtures — 
combining the overt motives with the new themes unique to the sym- 
bolic language. In these cases symbolic ambivalence is present. The 
imaginative themes cluster into two polar areas. One of these is 
usually stressed overtly and one is not. 

When this pattern of scores develops, a conflict is defined. The 
kind of conflict is determined by the location of the thematic clusters 
around the circle. The last case presentation illustrated a phobic- 
counterphobic conflict in which docile weakness covered underlying 
conflict between strength and weakness. 

When a patient's multilevel pattern of scores tends to cluster into 
two areas in the diagnostic circle we speak of the conflict axis. In 
the last two illustrations of phobic-counterphobic behavior the conflict 
axis was I- A. Other common dichotomous clusters are D-H which 
defines sado-masochistic conflict; O-K denoting nurturance-depend- 
ence; M-E denoting love-hate, etc. The relation between fantasy 
heroes and their protagonists often defines such reciprocal clusters. In 
one story the disappointed hero is rejected by the exploitive lover. In 
the next story the hero spurns his heartbroken mate to follow his own 
selfish goals. In the language of the notational system these are ex- 
pressed as: 

1 ) Hero Lover 

G C 

2) Hero Mate 

C G 

A conflict of rejecting versus being rejected is thus suggested. The 
conflict axis is C-G. The level of symbolism can help define the con- 
flict axis — the focal centers around which the patient's behavior at 
overt levels tends to cluster. Symbols can predict basic multilevel 

Sweet Symbols Behind a Facade of Sweetness. In the last few 
pages we have used some tentative case illustrations to fit together data 


from three levels of personality. These have served to underline some 
earlier principles and to introduce some new concepts (such as the 
conflict axis). We turn now to a third type of case which will con- 
tinue this process of review and preview. 

The subject in this case produced fantasies that are models of 
"sweetness and light." The heroes are affectionate, conventional, and 
generous. When evil or hatred appears it is met by virtue and agree- 
abihty — and usually transformed into good. All the endings are happy. 
In the case of the TAT even those cards which are loaded with nega- 
tive stimuli are transferred into positive. One card is seen by most 
people as portraying a girl slumped beside a gun. The case in ques- 
tion sees this as a girl overcome with joy, the vague gun-like object 
beside her becomes a "gift" from a loved one which has led to the 
joyous collapse. Another card which pictures a sprawled-out figure 
described by most people as dead or wounded is seen by our subject 
as a "sleeping man exhausted from a day of good work." Figure 21 
shows how the symbolic scores might profile. 

The scores for Level II-C self and father are also included, showing 
that the pollyanna pattern of optimism and goodness appears at the 
level of conscious description. Now it is quite possible to have siveet 
symbolic themes and hostile self-perceptions. When this occurs we 
have a conflict between a rough exterior and a symbolic heart of gold. 
The case profiled in Figure 2 1 shows no conflict, however. A sweet 
and loving fagade covers sweet and loving symbols. 

This means that the subject tells us — in the language of fantasy — 
the same message that he has expressed in conscious description. He 
denies hostility at Level II-C and both "layers" of Level III-T. We 
have assumed that a defensive process leads to extreme avoidance of 
negative affect in conscious description. The same process spills over 
into the level of imagination. We call this phenomenon a rigid in- 
varicmce. We conclude that the same anxiety process which imbalanced 
Level II-C in the direction of socially approved motives is at work at 
Level III-T. We suspect that hostility and assertive, bitter feelings 
are so anxiety provoking that they cannot be expressed in the more 
flexible language of symbolism. They cannot appear in responses to 
the cards on which the average person reports them. They cannot 
even be projected onto the fantasy "others." Even if the stimulus 
picture on the card pulls for hostile themes, this subject can maneuver 
his perceptions to avoid them. Ravaged corpses become sleeping 

Three points are worth stressing — the first two are familiar and the 
other is new. The first: fantasy themes give a rough indication of the 
interpersonal source of anxiety and the amount of it. If the themes 









Figure 21. Level II-C & Level III-T Hero & Other Profiles for a Rigidly Conven- 
tional Patient. 

that are strictly avoided at Levels I and II are also eschevv^ed at Level 
III we may speculate that these themes are the source of anxiety for 
this patient. The consistency of the avoidance at all levels indicates 
roughly how much anxiety. In the illustrated case we might guess 
that there is plenty of anxiety connected with the expression of ag- 
gression and antisocial motifs. 

Symbolic Themes Are Not Always 'Tre conscious'' 
or Opposed to Consciousness 
A second issue illustrated by this case concerns the nature of fantasy 
expressions. The fact that themes appear in dreams or projective tests 


does not mean that they are necessarily different from conscious per- 
ceptions. Autistic productions are not an automatic "pipehne" to the 
underlying motivation. They can simply repeat the pattern of secur- 
ity operations employed at the overt levels. 

This point has not been made clear in the psychological literature. 
There is a common tendency to assume that symbolic or projective 
data inevitably denote repressed or unconscious material. An amus- 
ing illustration of this fallacious assumption concerns the symbolic 
expressions of Nazi leaders who were given personality tests while 
awaiting trial in Nuremburg. These findings have been described in 
an excellent book by Dr. Gilbert, the prison psychologist ( 3 ) . One of 
those tested was Colonel Hoess, the S. S. official in charge of the 
Auschwitz concentration camp. It has been estimated that this man 
was directly responsible for the deaths of over two million prisoners. 
As such he probably ranks among the most murderous and sadistic 
human beings who have ever lived. A book reviewer for a psycho- 
logical journal noted w^ith surprise that the fantasy test expressions of 
Colonel Hoess were loaded with savage, cruel hostility. This led the 
reviewer to wonder why Colonel Hoess would have repressed sadism 
in his TAT stories when he overtly acted out so much aggression in 
his behavior. 

It seems clear that this psychologist was erroneously equating 
fantasy productions with repressed or unconscious material. The 
theory of variability developed by the Kaiser Foundation research 
(see Chapter 13) would expect that a person who expresses such in- 
tense hostility and who rigidly avoids tender, humanitarian feelings at 
Levels I and II would probably be unable to tolerate positive feelings 
even at the level of symbolism. Colonel Hoess, we suspect, was least 
anxious when he was employing cold, sadistic security operations. It 
is not surprising that the same avoidance of affiliative emotions re- 
appeared at Level III-T. Three-layer expressions of the same theme 
are typical of chronic maladjusted characters. 

Variation in Depth of Symbol Instruweiits 

The third issue evolving from this sample case concerns the defini- 
tion of levels and sublevels. At Level II we obtain the picture that the 
subject wishes to present to us — his conscious reports. At Level III he 
communicates in an indirect language that need not be bound by the 
limits of the real world. We have assumed thap Level III is related to 
the private world. x\ll human beings have a world of mental reserva- 
tions which are more or less distinct from what they directly express. 
The different layers of symbolic behavior, it seems, come closest to 
expressing these. 


But in many cases Level III hero and other scores are the same as 
Level IL The same themes are emphasized or avoided at both levels. 
The bland hysterical personality just presented was one such case. 
The sadistic Nazi executioner whose symbols were hostile is another 
such case. Still other patients stress pessimistic themes. They are de- 
pressed and bitter at Level II and an equally unhappy blackness stains 
their symbols. We surmise that trustful, tender emotions are so threat- 
ening to these patients that they must avoid them even in fantasies. 

When the autisms of Level III are the same as the conscious descrip- 
tions of Level II, a puzzHng question occurs. In this event it would 
seem that the private language is the same as the overt and conscious. 
This is faintly paradoxical. A confusion of levels is suggested. Since 
the private is also public, it suggests either that these patients have no 
thoughts that are exclusively private (i.e., secret) or that the symbols 
have failed to express the private. If the first conjecture is true, then 
the subject is unusually frank and honest — having, as it were, no men- 
tal reservation. If the second is true, then the subject is unusually 
repressive and secretive — he succeeds in blanketing his symbols with 
the same avoidance tactics that characterize his conscious expressions. 
The former would be bluntness and insight. The latter would be 
evasion and symbolic denial. 

According to the theory developed in this chapter, if symbols dupli- 
cate the extreme imbalances of the conscious level, then considerable 
information about the patient's anxiety system is available. The sub- 
ject compulsively avoids hostihty in consciousness. Even in symbol 
he cannot tolerate an expression of the negative. If he does have 
private feelings of bitterness or aggression, they are not allowed sym- 
bolic expression. He does not dare let his symbols express his private 
feehngs. Level III instruments (e.g., TAT tests) are, in this case, tap- 
ping only Level II. They do not "dig down deep enough." A very 
thick layer of defensive avoidance exists — so that the Level III instru- 
ments fail to get at the private world. 

For some patients in some situations projective tests such as the 
TAT fail to reach anything different from conscious report. Consider 
a patient who covers feelings of distrust and depriv^ation with a facade 
of extroverted congeniality. If he takes a TAT in connection with ap- 
plying for a desirable job, the underlying feelings may not appear in 
his fantasy stories. The same patient applying for therapeutic help in 
the psychiatric clinic may produce TAT stories which express his 
feelings of sorrow and defeat. The same person, were he attempting 
to "buck for a medical discharge" from the Army, might overexag- 
ger*ate his depressive feelings on the TAT — they might even be picked 
up by Level II instruments. 


The fact that we obtain Level III fantasy protocols does not mean 
that we are necessarily tapping the private world of the patient. Meth- 
ods of measuring fantasy vary in depth. The "preconscious" self in 
some subjects is closer to consciousness than the images of symbolic 
others. Dreams seem to produce themes which are most distant from 
conscious report and thus deeper. To use Freudian terminology, some 
parts of some dreams seem to tap the primary processes characteristic 
of schizophrenic or infantile thought. Projective tests and fantasies 
are generally "preconscious" and probably tap secondary processes. 

Level III instruments might be compared with drilling machines 
which tap geological strata. We cannot assume that the TAT or a 
fantasy automatically taps private feelings. The depth and thickness 
of the strata of conscious report is a crucial and variable factor. As il- 
lustrated in Figure 22, a flexible person with minimum anxiety (Case 
2) may confide his "preconscious" feelings in interviews. He has some 
conscious awareness of his ambivalences and is able to discuss them. 
When the conflict is more severe and anxiety greater (Case 3) more 
indirect instruments may be required to hit the private "layers." This 
type of patient may present the same picture in a check list and even 
in therapy interviews. The TAT themes and dreams may indicate the 
other side of the ambivalence. In other cases (Case 4) the anxiety ac- 
companying certain interpersonal emotions is so great that they do 
not appear in the expressions of the "preconscious" self. The absent 
themes may be projected on the "preconscious" world or they may 
appear only in the subliminal expressions of Level IV, 

Determining the Depth of the Measuring Instrument 

Use of a Level III test thus does not guarantee that the "precon- 
scious" will be discovered. This fact does not in any way lessen the 
value of these tests. They always assist in determining the rigidity and 
depth of the defensive processes. A vital part of interpersonal diag- 
nosis is to determine the amount of anxiety and the way it operates to 
inhibit or deny certain touchy emotions. 

The interpretation of a projective test or dream is greatly facili- 
tated if we know the depth of the private world and the "thickness" 
of the conscious defensive processes. These are determined in two 

As soon as the data are obtained from a Level III test we compare 
them with the data from Level II. If the symbols clearly duplicate the 
rigidities of Level II, then we estimate the conscious defensive pro- 
cesses extended down to the depth of the test. The more rigid the 
similarity between the symbolic instrument and Level II — the greater 
the anxiety, the thicker the defensive strata. In these cases the thematic 





at all 


^1 , §^ 

1 .' 



f " s s 


















II" i 







ex a. 













S A E ^ 








aspects of Level III add nothing new. As each additional depth test 
repeats the same themes, we learn nothing new about -what themes are 
present in the elusive warded-off private world — but we do learn 
something about how deep it is and how much anxiety is tied to it. 
The first way of determining the depth of the measuring instrument is 
to compare its themes with Level IL The more discrepancy, the 
further from consciousness. 

A second method for determining the depth of the measuring rod 
involves use of internal cues of defensiveness. The two best internal 
cues for estimating defensiveness from projective tests are mispercep- 
tions of stimuli and avoidance of specific themes. These issues are dis- 
cussed in the next chapter. 

Situational Relativity of Symbols 

Another factor which exerts strong pressure on symbolic expres- 
sions concerns the motivation of the patient in the particular situation. 
Consider a subject whose private feelings concern weakness and des- 
pair. Suppose he is applying for a job which entails executive responsi- 
bility, and as part of the application procedures is administered a 
fantasy test. Since the job situation would tend to motivate the ex- 
pression of strength and assertion — the expression of his private feel- 
ings would be threatening. The feelings of inferiority and depression 
would, therefore, not be likely to appear. The subject might really 
have the wishful fantasy of retreating from the demands of the world 
into an isolated, lonely, passive life. He might secretly yearn to be a 
forest-fire watcher or a beachcomber. But if the employment inter- 
viewer asks him projective questions about his hopes for himself, he 
might respond with the wishful statement: "I want a responsible, 
managerial job with a big company." We are, in this last statement, 
obviously not tapping the symbolic level. Level III is being used to 
support the overt presentation. 

The cultural situation in this example was the "job application." 
The cultural situation we are concerned with in this book is the 
psychiatric clinic. We are attempting to develop a functional diag- 
nostic system which will lead to predictions about clinic behavior. 
When patients come to a psychiatric clinic for diagnostic evaluation 
they vary considerably in their motivation. Their symbolic produc- 
tions will vary considerably. Many patientis sense psychological 
evaluation and therapy as a threat to their imbalanced, inflexible ad- 
justments. These patients may avoid in their fantasy tests the same 
themes that cause them anxiety at Levels II and I. Motivation of the 
patient, as well as the source and amount of anxiety, is a complicating 
factor in symbolic interpretation. 


This is not a particularly distressing complication to the psychol- 
ogist evaluating the patient. Certainly we concede that motivation 
varies from patient to patient and these variances influence the Level 
III material. But "amount of motivation," far from being a distracting 
irrelevant factor, is actually most central to prognosis. If motivation 
effects the production of symbols, then symbols can help estimate the 
amount and kind of motivation. They can help plan the correct 
therapeutic program that works with and does not clash headlong 
against the unique defensive set-up of the particular patient. 

Let us recall the hysterical patient (page 181) who presented sweet 
symbols behind a facade of sweetness. The TAT themes duplicated 
the conventional, pious traits of Level II. This suggests that the de- 
fensive structure is "thick" — the anxiety accompanying antisocial or 
negative feelings is very high. The patient might be panicked by being 
referred to psychotherapy. He might react with a severe anxiety at- 
tack. Most likely, he would react by increasing the bland denial of 
pathology. Both of these reactions would postpone the onset of effec- 
tive psychological help. 

At the Kaiser Foundation Psychiatric Clinic a patient with such a 
hysterical "normality syndrome" would not be rushed into psycho- 
therapy. The nature of treatment might be explained to him. The 
intake worker might discuss with the patient (in nontechnical lan- 
guage) the nature of his current adjustment — by referring to the pa- 
tient's own claims to health and hypernormality. This is done sup- 
portively. To use psychoanalytic terminology, it is done "from the 
side of the ego." The rigid claiming of goodness is not attacked as a 
defense but might be praised as a valuable means of adjustment. No 
speculation is made about underlying motivation. The advantage of 
living with the present adjustment (and the symptoms it involves) 
might be discussed. The function of psychotherapy as possibly lead- 
ing to different solutions might be mentioned, but not pushed. The 
patient is offered the opportunity to return to the chnic at any later 
date if symptoms worsen or if he feels it worthwhile to learn more 
about his patterns of living and the possibility of changing them. 

This process might be called "planting the seed." The TAT, in this 
case, provides the information that, at present, this patient is deeply 
committed to hysterical bland techniques of adjustment. Even in 
fantasy these motives appear. The TAT predicts that exploration of 
other feelings is, at present, not likely. It predicts that the patient in 
the immediate future is moving away from and not toward his 
warded-off feelings. 

Level III thus has many sublevels. Some of these are determined by 
the nature of the measuring instrument. Dreams appear to be the 


i-J !-? 

h i 


SqOv — O 5'000 QC30 

{EuuoiuadXH > 2: 2 iQ 2 
-ajqisuodsa-a < "• ^ "" '^ 

g I 

2 ;3 



1; — - 

— Q— rM-H ,-1 ru '-• '" r-j—*'— 

tj H onsisspjBf^ Cj '^ 00 


ft- C 

R O '5 '2 '"' .=; " « 



deepest sublevel. Wishful fantasies and projective tests are less deep. 
These sublevels vary from situation to situation. In six months the 
patient described here may return to the clinic. He may see it this 
time, not as a threatening institution ready to expose his antisocial 
feelings. He may be less defensive. His thick protective strata (which 
are measured by the extent to which Level II themes penetrate and 
duplicate Level III data) may be considerably diminished. His sec- 
ond testing battery may show the same conscious description but a 
more changeable TAT. He would then be considered more ready to 
deal with his conflict between bland sweetness and the underlying feel- 

Incidence of Level lll-T Behavior in Various Cultural Samples 

Detailed summaries of the research findings involving Level III 
behavior are presented in the clinical chapters and Appendix 3. At this 
point the percentage of Level III-T types found in several sympto- 
matic and institutional samples are presented in Table 5. 

It will be observed that psychotics manifest "preconscious" sadism 
(DE) and distrust (FG); medical controls (normals), underlying 
power (AP) and narcissism (BC); obese patients, power (AP) nar- 
cissism, and hostility; ulcer patients, an intense amount of underlying 
dependence (JK), etc. 

Significance tests and discussion of the clinical and theoretical im- 
plications are presented in Chapter 24. 


1. Blum, G. S. The Blacky Pictures: a technique for the exploration of personality 
dynamics. New York: The Psychological Corporation, 1950. 

2. Freud, S. The basic writings of Sigmund Freud. New York: Modern Library, 
1938. Copyright, New York: The Macmillan Co. 

3. Gilbert, G. M. Psychology of dictatorship. New York: The Ronald Press Co., 

4. Iflund, B. Selective recall of meaningful materials as related to psychoanalytic 
formulations in certain psychiatric syndromes. Unpublished doctor's dissertation. 
University of California, Berkeley, 1953. 

5. Kluckhohn, C, and H. A. Murray. Personality in nature, society and culture. 
New York: Alfred A. Knopf, Inc., 1949. 

6. Kris, E. On preconscious mental processes. Psychanal. Quart., 1950, 19, pp. 540-60. 

7. Langer, S. Philosophy in a new key. Cambridge: Harvard University Press, 1942. 

8. Murray, H. A. Thematic Apperception Test. Cambridge: Harvard University 
Press, 1943. 

9. Powelson, D., and R. Bendix. Psychiatry in prison. Psychiat., 1951, 14, 73-86. 

10. Thurber, J. "The Secret Life of Walter Mitty." Copyright, 1939, James Thurber. 
Originally published in The New Yorker. 

11. ToMKiNs, S. S. The Thematic Apperception Test. New York: Grune & Strat- 
ton, 1947. 


The Level of the Unexpressed: 
Significant Omissions' 

The levels of personality described in the last three chapters have 
moved steadily from overt observable behavior (Level I) through 
conscious description (Level II) into the two private or underlying 
layers of the "preconscious" (Level III Hero and Other). This se- 
quential progression frpm the external to the internal brings us now 
to the deepest level of personality — Level IV. 

This is called the level of the unexpressed. It comprises those inter- 
personal themes which the patient consistently, significantly, and spe- 
cifically omits in the three other levels. 

The Two Criteria for Defining Level IV Themes 

The essence of Level IV themes is that they not be expressed in ac- 
tion, in consciousness, nor in the "preconscious." The first criterion is 
that the themes be avoided at these levels. This negative evidence can- 
not in itself be taken as proof that the themes are "dynamically" ab- 
sent nor that they exist in deeper strata of the personality. To accept 
this absence at one level as a sign of presence at another level is to com- 
mit the ancient fallacy of reversal which was discussed in the preced- 
ing chapter. 

The second criterion for defining Level IV requires evidence that 
the themes are actively avoided. It must be demonstrated that the sub- 
ject selectively and stubbornly refuses to respond to these themes 
when they are appropriate in the situation. It is not enough to report 
that a patient fails to express a particular cluster of themes — let us say 
competitive hostility — at the three top levels of behavior. In addition, 

* This level of personality has not been studied systematically by the Kaiser Foun- 
dation psychology research group. The definition and discussion in this chapter is 
tentative and suggestive. Readers who are interested in the current clinical or re- 
search applications of the interpersonal system can safely omit this chapter. 




it must be shown that he has been exposed to situations in which he is 
naturally or consensually expected to perceive, react to, or express 
these themes and that he has refused to do so. 

This level of personality has received little empirical attention and 
is therefore not included in the systematic or clinical studies described 
in this book. This chapter will present a definition and a survey of 
some tentative, unvalidated techniques for measuring Level IV. Al- 
though in the subsequent chapters no reference will be made to Level 
IV, the present discussion is included as a preliminary description of 
this incomplete aspect of the interpersonal system. 

From the functional point of view the existence of unexpressed 
interpersonal themes seems to be of some importance. It might be 
argued that in clinical practice we are interested, not in the absent 
motives, but in the strata of ego functions which lie above them and 
which seem to be organized in warding them off. When we deal with 
a patient who presents conventional, bland themes at Levels I, II, and 
III, our diagnostic attention is obviously going to be focused on these 
ego processes. On the other hand, it is useful to know that inter- 
personal themes comprise the Level IV significant omissions. These 
themes can be expected to be anxiety-laden. If the patient is con- 
fronted with them, panic may develop. Level IV defines the "touchy 
spots" most vigorously and desperately avoided. 

In the Kaiser Foundation research we have tended to concentrate 
on the three more overt levels. In psychoanalytic language we have 
been attempting to develop an ego psychology. Research is now being 
planned which will investigate some of these aspects of the signifi- 
cantly omitted. 

The methods being considered for these future studies will now be 

The Measurement of Unexpressed Themes 

It has been pointed out that there are two empirical criteria for the 
measurement of Level IV themes. They must negatively be demon- 
strated to be significantly absent at the three top levels; they must 
positively be demonstrated to be actively avoided. Two separate 
measures of Level IV are therefore available — the omission and the 
avoidance scores. These two criteria will be treated separately. 

The Measurement of Omission of Interpersonal Themes. 
The first criterion is easily measured. The scores at Levels I, II, III 
Hero, and III Other are examined to see what behaviors are consist- 
ently avoided. This can be done by inspection of the profiles or by 
means of arithmetical indices. 



Figure 23 presents the profiles of an illustrative patient who has 
clearly failed to express rebellion or unconventional themes. 







Figure 23. Profiles of Interpersonal Behavior at Four Top Layers of Personality 
Illustrating the Avoidance of Rebellious (FG) Behavior. 

By inspection we see that the FG and DE octants are conspicuously 
neglected. The Level I profile is obtained from sociometric ratings of 
other group therapy patients. It will be observed that the fellow group 
members did not use any FG or DE words to describe this patient's 
impact upon them. In his Level II-C self description he completely 
denied any of these behaviors. Even his fantasy heroes and others 
fail to receive any of these motivations. It is possible to determine 
by glancing at these four interpersonal profiles that there is a con- 



sistent tendency to avoid the expression of resentful, rebellious, hostile 

In addition to the diagnosis of Level IV omission by inspection of 
the profiles it is also possible to employ mathematical techniques. 
These allow the Level IV omission profile to be plotted in terms of a 
single summary point. These are determined by the horizontal and 
vertical indices just as in the case of the three overt levels. 

There are many ways in which Level IV omission scores can be 
calculated. It is possible to study at each of the four more overt levels 
the sectors which the subject neglects. We could study all of the items 
on the interpersonal check list not used by the fellow group members 
in rating the patient. These could be treated hke the "yes" scores, 
converted into the horizontal and vertical factors and plotted. This 
provides an omission-at-Level-I score. The same could be done for 
the interpersonal check list self-description at Level II. 

We want, however, a single summary score which will give the 
over-all pattern of what is omitted at Levels I, II, and III. A single 
way of estimating the Level IV omission score is to take the standard 
score indices for Levels I, II, III Hero, and III Other (which are used 
to plot the diagnostic people) and to establish the mean horizontal 
and vertical indices for these four scores. 

For the patient profiled in Figure 23 the scores were: 







Level I-S 
Level II-C 
Level III-T Hero 
Level III-T Other 






Mean = ^ = 57 

f = " 

These two mean indices provide a summary of the themes the pa- 
tient has expressed at these four layers of personality. If these indices 
are each subtracted from 1 00 they give a summary of what the patient 
has avoided at Levels I, II, and III. The mean omission scores for this 
patient are vertical =43, horizontal =21 (see Figure 24). This is 
located in the FG octant and indicates that this patient compulsively 
and markedly omits rebellious, unconventional, and bitter behavior in 
his expressions at Levels I, II, III Hero, and III Other. 

The Measurement of Significant Avoidance of Interpersonal 
Themes. There are two criteria for determining significant omis- 


























— -V 





, 20 














\ ' 











Figure 24. Diagnostic Grid Containing Summary Points for the Four Top Layers 
of Personality and the Level IV Omission Score Calculated by Subtraction. Key: 
Radius of diagnostic grid = 6 sigmas, / =: Level I-S; S =r Level II-C description of self; 
H = Level III-T (Hero): O = Level III-T (Other); IV = Level IV omission score. 

sions at Level IV. The first is to demonstrate that certain themes are 
consistently omitted at the upper levels. A method has just been pre- 
sented for determining this type of Level IV score. 

The second criterion involves the demonstration that these themes 
are actively avoided by the patient at the upper levels. The omissioii 
score just calculated for Level IV indicates that the patient has 
"claimed" or "expressed," and by a process of subtraction we obtain 
the Level IV index. The avoidance score is obtained in a different 

There are at least three techniques for estimating the tendency to 
avoid themes at Levels I, II, and III. The first of these is based on sta- 
tistical procedures and estimates the tendency to fail to perceive or 


express interpersonal themes where, consensually, most others do per- 
ceive or express them. 

It is possible to determine for each interpersonal test item at each 
level the probability of its being expressed. The interpersonal check 
list employed at Levels I and II has been subjected to intensive sta- 
tistical analysis. The percentage of the clinic sample expected to re- 
spond to any particular word has been determined. The 128 words 
on the check list have been classified into four groups along an inten- 
sity dimension in terms of its probabiUty of occurrence. Intensity 1 
includes words which are checked as "true-about-self" by approxi- 
mately 90 per cent of psychiatric clinic patients. There are minimal 
amounts of the trait which almost everyone is willing to attribute to 
himself. Intensity 2 includes words employed by 50 to 90 per cent of 
the clinic population. Intensity 3 employs more intense themes used 
by relatively fewer (10-50 per cent) patients. Intensity 4 employs ex- 
treme loadings of the theme which are rarely checked by clinic pa- 
tients. Here are illustrations of check list items at the four intensities 
for the interpersonal variable D, which includes themes ranging from 
appropriate sternness to punitive sadism. 

Intensity 1 (very common, expected): Able to be strict 

Intensity 2 (average-moderate): Stern but fair 

Intensity 3 (fairly intense): Sarcastic 

Intensity 4 (rare-extreme): Cruel and unkind 

Each item on the check list is thus weighted in terms of its consensual 
or average usage. If a patient does not check the Intensity 1 word for 
any interpersonal variable, he is failing to attribute to himself a mild 
amount of this theme which 90 per cent of the clinic population does 
express. In determining the Level II diagnostic indices, the weightings 
of the items are not considered. All the words used in every octant are 
fed into the formulas. 

To determine the significant avoidance score these weights are of 
usefulness. At Levels I and II which employ the check list we can 
study the pattern of avoidance of Intensity 1 and 2 items. Weights can 
be assigned so that the failure to check these mild, average, statistically 
common items about the self can be cast into numerical indices of sig- 
nificant avoidance. 

At Level III (as measured by the TAT) the test stimuli are pictures 
which portray human beings in interaction. The probability of any 
given interpersonal theme being expressed in reaction to any particular 
card has been determined. Thus we discover that 85 per cent of the 
clinic population respond with the fantasy theme HI (sorrow or 


guilt) on card 3BM, whereas less than 2 per cent will produce a story 
involving AP (dominant power). This information makes it possible 
to weigh the thematic pull of every card on the TAT (or any other 
fantasy test). Patients who consistently resist the card pull for any 
particular interpersonal theme can be assigned Level IV avoidance 
indices which are based on the percentage expectance. 

This general method for measuring significant omission has been 
previously described by William E. Henry in his monograph on TAT 
Analysis (1). He defines two areas of TAT content interpretation: 
positive content and negative content. "The difference between 
positive and negative content can be summarized in this way: the 
analysis of positive content is concerned with what the subject actu- 
ally has said, the analysis of negative content is concerned with what 
the subject has failed to say and with what he might have been ex- 
pected to say considering the usual responses made to that picture." 

Normative and validating research on the significant avoidance 
scores based on these statistical characteristics of the tests for Levels I, 
II, and III is now being done. Since these studies have not been com- 
pleted, the Level IV avoidance scores are not included in this volume. 

There are two additional sources of Level IV significant avoidance 
data which will now be briefly reviewed. Both of these are based on 
special psychological techniques for measuring the selective factors in 
memory and perception. 

A test of repression developed by Boris Iflund (2) seems to pro- 
vide an excellent measurement of Level IV data. The Iflund test deals 
with selective memory factors. It consists of 34 cards, each of which 
contains a picture. Twenty-eight of these are illustrations of personal- 
ity needs (as listed by Murray). Six pictures are bu^er cards which 
contain nonloaded (street or landscape) scenes. The subject is told 
that he will be shown the stack of cards, each card being exposed for 
5 seconds. He is told that after the entire deck has been shown to him 
he will be asked to recall as many as he can. After the subject has re- 
called as many as he can the nonbuffer (i.e., thematic) cards which he 
recalled are removed from the deck. The same process of presenta- 
tion and recall is repeated until all cards have been recalled. In indi- 
vidual administration an inquiry period after the test is employed to 
clarify accuracy or misperceptions of the cards. 

This test is based on the psychoanalytic theory of repression. It is 
held that the themes which the subject remembers last are subject to 
more repression than those he recalls first. The serial order of recall 
is believed to relate to intrapsychic defensive processes holding sensi- 
tive themes for awareness. 


To the extent that the Iflund test does isolate warded-ofF themes, 
it can be used as an estimate of the stratification of personality and as 
an indicator of Level IV themes. The most-forgotten themes, i.e., 
those remembered last, should be close to the Level IV omission score. 

Research on the Iflund test is currently uncompleted and is not 
included in the systematic and diagnostic studies reported in this book. 

A third technique for determining Level IV significant avoidance 
scores has been suggested. This involves the theory of perceptual 
vigilance or perceptual defense in relation to sensitive or warded-off 
emotional stimuli. Methods have been developed by experimental 
psychologists for determining the level of perceptual readiness to re- 
spond to varied stimuli. The tachistoscope (which is a machine for 
exposing stimuli cards to an observer at split-second speeds) is em- 
ployed in these experiments. There is some evidence suggesting that 
the speed of recognition varies in relation to the emotional loading of 
the stimuli. Subjects whose personalities are mobilized to ward off 
hostility from awareness tend to require slower speeds of presentation 
in order to perceive hostile motifs on cards. 

To the extent that this theory of motivated perception and per- 
ceptual defense holds true, tachistoscope recognition speed becomes an 
estimation of Level IV. The themes which are recognized most 
quickly should be those which are expressed or consciously claimed at 
the upper levels. Those which are recognized at the slowest tachisto- 
scope speeds should also be avoided at the levels of action and con- 
scious report. Perceptual defense thus becomes an estimate of the 
significantly avoided and a measure of Level IV. 


1. Henry, W. E. The Thematic Apperception Test technique in the study of culture- 
personahty relations. Genet. Psychol. Monogr., 1947, 35, 3-135. 

2. Iflund, B. Selective recall of meaningful materials as related to psychoanalytic 
formulations in certain psychiatric syndromes. Unpublished doctor's dissertation. 
University of California, Berkeley, 1953. 


The Level of Values: The Ego Ideal 

There Is another area of human behavior which because of its func- 
tional value and theoretical uniqueness has been designated as one of 
the operating levels of personality. This is Level V — the level of 
values. It includes the interpersonal aspects: ideals held by the indi- 
vidual — his conceptions of "lightness," "goodness," of w^hat he should 
like to be. 

The last four chapters have presented Levels I through IV, working 
sequentially from the public overt aspects of behavior into the more 
private, unexpressed areas. At this point the trend is reversed. The 
numerical designation of Level V suggests that this is the deepest 
level — which, of course, it is not. Level V is concerned with con- 
sciously reported ideals. The subject is asked to list, or describe, or 
check his picture of how he should like to be. 

Level V Is aji Independent Area of Fersonality 

Level V, as presently measured in the interpersonal system, is not 
a very complicated or deep measurement. It simply gives us a pic- 
ture of how the subject wants us to see his ideals. It tells us which 
values he consciously stresses. The subject may privately have dif- 
ferent goals and stress different feehngs. His private value system 
may be in contradiction to his openly reported principles. 

The working procedures of the Kaiser Foundation research proj- 
ect simplify this complexity of values. The general empirical ap- 
proach of this system of personahty is to select several narrow opera- 
tionally defined areas of behavior (which are called levels) and to 
utihze the same matrix of interpersonal variables to measure behavior 
at these levels. The levels are defined by the nature of the data, that is, 
by the way it is obtained, by the technical context of the measurement 
process. If the subject produces fantasy themes in response to pro- 
jective test stimuli then Level III is defined. When he attributes inter- 
personal themes to his ideal then Level V is defined. 


Although Level V is a rather simple measure of consciously re- 
ported values, it possesses a clear-cut statistical independence. That is 
to say, it does not duplicate the other levels. Patients' descriptions of 
their ideals are very often quite different from their conscious self- 
descriptions and their fantasy expressions. This measure seems to vary 
independently of the other levels. This offers reasons for expecting 
that it may serve a unique psychological function and possess a unique 
clinical application. 

The Universality of Value Systems 

The Level V value system gives us a picture of the interpersonal 
standards which the subject holds for himself. The notion of ideal, it 
should be noted, is widely accepted and natural. During our testing 
procedures patients readily take to the task of describing their ideal. 
The concepts of "right" and "good" and the interpersonal themes 
associated with these values seem to be taken for granted. 

The vital and universal process of idealization has been recognized 
by almost every personality theorist. Kluckhohn and iMurray ( 1 , p. 2 1 ) 
for example, state that: "One of the important establishments of a per- 
sonality is the ideal self, an integrate of images which portrays the 
person 'at his future best,' realizing all his ambitions." The related 
concepts of ideal, superego, and introjection have been receiving in- 
creasing emphasis in psychoanalytic theory. 

Ethical standards appear to exist in all cultures. In an earlier chap- 
ter the universality of symboHc behavior was noted. The same can be 
said for moral standards of conduct. It is hard to conceive of a society 
or a social group which does not possess many principles of "rightness" 
and "wrongness." Not all the members of a society necessarily share 
(publicly or privately) the same set of values. Some may assign posi- 
tive values to force, others to deceit, others to charity. In many so- 
cieties women may publicly accept standards which are different from 

Although the specific qualities to which "goodness" and "bad- 
ness" are attached may vary, what seems to remain constant is the as- 
sumption that there are standards of right and wrong. 

It is these principles which we tap in Level V when we study the 
interpersonal themes which the subject idealizes and those which he 
avoids attributing to his ideal. 

The Function of Value Systems 

This poses the questions: Why do individuals develop standards 
and ethical principles of behavior? What is the function of these 


A detailed discussion of this topic is well beyond the scope of this 
book, encompassing as it does the genesis and meaning of ethics and 
morals. The general assumption about human motivation employed 
in this book does suggest certain approaches to these questions. 

The basic function of the individual's interpersonal behavior is to 
ward off survival anxiety. Any personality pattern can be viewed as 
an attempt to come to terms with the social environment. In this light 
the development and maintenance of value systems can be seen as pro- 
viding several bulwarks against anxiety. 

First of all, the acceptance of certain ideals tends to link the indi- 
vidual to strong forces in his world. By taking over and expressing 
these ideals the subject identifies himself with powerful images of 
rightness. The standards may come from his parents, from his con- 
ception of religious figures, from the standards held by his social 
groups, etc. Generally they come from all these sources. Symonds 
(2) has described the process by which the individual combats his 
feelings of weakness and develops a feeling of omnipotence by taking 
as models and values those of the group: "One looks for support by 
acceding to the wishes of society through its laws and customs, so 
that one feels secure as a member of the group and derives power from 
the group. A still further development is to align oneself with the 
universe and to look to God for strength. So the religious person, by 
obeying the rules of morality, is continuing this process of gaining 
strength for himself by aligning himself with superior forces." 

By taking on standards and ideals the individual wins approval and 
attempts to ward off disapproval. Heightened self-esteem and the 
avoidance of shame and inferiority can be achieved by the acceptance 
and expression of value systems. It appears that all human beings 
maintain this one unique area of their personality which reflects their 
conception of what they should or could be. 

Like behavior at other levels, the value system may play a de- 
structive and unsettling role in the total personality structure. Ideals 
which are too elevated or standards which are too strict may lead to 
severe conflict with other levels. Thus the Level II self-conception 
may fall far short of a rigid, demanding set of ideals — with a resulting 
feeling of guilt and self-dissatisfaction. This phenomenon has been 
noted at other levels where extreme, exaggerated behavior at one level 
creates new circles of conflict and anxiety. 

Human beings presumably develop standards in an attempt to lessen 
anxiety, win approval, or to win security through linkage with power- 
ful parental and societal forces. But the complexity of social adjust- 
ment generally creates the tragic paradox of humaa nature — the tech- 


niques for avoiding anxiety at one level are related to the activities of 
other levels. They cause tension at other levels w^hich in turn may 
increase anxiety and lead to an increase in the original behavior. 

Kluckhohn and Murray ( 1 ) have commented on certain aspects of 
this process. They point out the relationships of aspirations and ideals 
to the "frustration and dissatisfaction" of overt behavior. "High as- 
pirations can cause unhappiness and discontent, while the process of 
low^ering aspirations to realizable levels is functional." 

This process of relaxing standards is, however, not a simple or 
voluntary procedure. The ideals held by individuals, like the behavior 
expressed at any other level, are not easily changed. This is, we pre- 
sume, because they play a vital functional role in the total personality 
pattern. Human beings develop ideals for the very important purpose 
of warding off survival anxiety and avoiding shame, weakness, and 
disapproval. These ideals develop and are expressed in reciprocal 
response to the activities of other levels. Frustrations (and the ac- 
companying anxiety produced) at Level I may result in a lowering of 
the associated Level V ideals; or they may result in an increase. Inter- 
personal behavior at any level of personality has the function of ward- 
ing off anxiety. This behavior can shift in response to stimuli from the 
external environment or in relationship to pressures or changes from 
other levels of personality. A most complex chain of multilevel 
processes is involved in any interpersonal pattern at any single level. 
The interpersonal ideals which we measure at Level V are not excep- 
tions to this principle. In the latter sections of this chapter some of 
these multilevel relationships (the indices of self-acceptance and ideali- 
zation) which involve Level V and the other establishments of the per- 
sonality will be reviewed. 

The Measurement of Interpersonal Ideals 

To obtain measurements of Level V behavior it is necessary to have 
the subject communicate his system of values. His ideals are then 
categorized in terms of the continuum of sixteen interpersonal var- 
iables. These scores are then treated in the same way as the scores from 
other levels; they can be formalized, standardized, diagramed, and 
then related to the total multilevel pattern. 

There are many methods for obtaining the raw protocol data for 
Level V. The subject can be asked to describe his ideals either in 
interview or in essay form. He can be given check lists or question- 
naires about his values. Regardless of how the data are collected the 
rating procedure is the same. The interpersonal aspects of these ex- 
pressions are coded into the language of the interpersonal system. 



At the present time the Kaiser Foundation project is employing 
three methods for obtaining Level V data. 

Scores from the interpersonal adjective check list on which the pa- 
tient rates his ego ideal are coded Level V-C. 

Ratings by trained personnel of the subject's ideals as expressed in 
diagnostic interviews are coded Level V-Di; in therapy interviews, 
Level V-Ti. 

The Kaiser Foundation research project routinely obtains Level V 
protocols (along with measures at seven other areas or levels) as part 
of the personality test battery. The key measuring instrument in this 
process is the interpersonal check list. Each patient uses this check 
list to rate first himself then his parents, his spouse, and his ego ideal. 
The instructions for the Level V-C test request the patient to check 
the items which describe "his ideal, his picture of himself as he should 
like to be." The patient is thus allowed to describe his value system on 
the same measuring instrument which he has employed to describe 
himself and three important family members. The empirical ad- 
vantages of this procedure for comparing behavior at different levels 
has been previously discussed. 

Figure 25 presents the Level V-C profile of a patient tested in the 
Kaiser Foundation clinic. For comparison we have also included the 



Figure 25. Illustrations of Level II-C and Level V-C Profiles for a Docile Patient 
Whose Ego Ideal Involves Strength. 

Level II-C self profile. These diagrams indicate that the patient sees 
himself as a weak, docile person. His ego ideal stresses themes of 
strength and power. 


The Functional Value of Level V 

Level V is a simple but useful diagnostic tool. Its first and most 
obvious application is the insight it gives us into the subject's value 
system. Human beings vary in the interpersonal themes they idealize. 
Some stress congeniality and conventional agreeability. Others em- 
phasize strength and assertion. Some prize competition. Others are 
concerned with frank, blunt honesty, or modest reserve. There seem 
to be relationships between diagnostic types and the nature of the 
ego ideal. Patients who stress dominance or submission at Level I-M 
emphasize pure power in their ego ideal. Patients who manifest either 
hostility or love at Level I-M are significantly less concerned with 
power and more involved with friendliness in their ideals. In other 
words, subjects whose actions fall along the vertical axis place their 
ideals at the top of the vertical axis (dominance). Subjects whose ac- 
tions locate on the horizontal axis place their ideals close to the hori- 
zontal axis (affection). 

Another and perhaps the most important use of the Level V score 
is obtained by comparing it with other levels. The discrepancy be- 
tween ideal and Level II Self provides an index of self -acceptance. 
This variable plays a most crucial role in arousing motivation for 
therapy. Similarly, the kind and amount of discrepancy between the 
ideal and the conscious descriptions of family members provides an- 
other set of valuable indices. These discrepancies are called the in- 
dices of idealization. 

The essence of these and the other interlevel discrepancies between 
the ego ideal and the other areas of personality is as follows: once we 
systematically locate the subject's ego ideal we can compare all the 
measures from the other levels of personality to see how close they 
fall to the ego ideal. Different theoretical and clinical implications 
are attached to these idealization indices. 

Limitations of the Level V Score 

The measurement of ego ideal employing the interpersonal adjec- 
tive check list is somewhat limited because of the tendency of all pa- 
tients to stereotype their ideals. In one sample of 207 routine clinic 
intake patients, 53 per cent placed their ego ideal in the managerial 
octant and 37 per cent in the responsible-hypernormal octant. This 
means that 90 per cent of all patients had ego ideals In the upper right- 
hand quadrant and less than 2 per cent placed their ego ideal in the 
lower (weak) half of the diagnostic circle. 

This homogeneity is a cultural stereotype. We might expect that 
some other cultures would stress aggression and some (oriental, for 


example) might idealize the passivity and modesty which American 
urban subjects so dramatically devaluate. 


1. Kluckhohn, C, and H. AIurray (eds.). Personality in nature, society and culture. 
New York: Alfred A. Knopf, Inc., 1949. 

2. Symonds, p. The dynamics of human adjustment. New York: Appleton-Century- 
Crofts, Inc., 1946. 


A System of Interpersonal Diagnosis' 

We have suggested in Chapter 6 that functional diagnosis of personal- 
ity involves two basic dimensions — interpersonal behavior and varia- 
bility. The preceding five chapters of this section have dealt with the 
concepts and measurement techniques by which we systematize the 
interpersonal dimension of personality. We have defined the inter- 
personal variables by which security operations of the human being 
can be classified. We have considered the levels at which this behavior 
is observed. 

With these theories and methods as background, it is now possible 
to consider the application of this personality system for interpersonal 

Purpose of Personality Diagnosis 

An examination of psychiatric nosology reveals considerable varia- 
tion in terms. Some diagnostic categories refer to the symptomatic 
picture. Some are clearly moralistic or evaluative epithets, e.g., in- 
adequate personality. Some refer to character traits, e.g., obsessive- 
compulsive. Some are global terms for disease entities which sum- 
marize many specific factors, e.g., schizophrenic and hysteric. 

In order to evaluate these terms it is necessary to inquire into the 
purpose of personahty diagnosis. Just why do we need a nosology? 
Just what is it to be used for? 

Psychiatric or personality diagnosis terms actually have many dif- 
fering uses, depending on the institutional or cultural context. In the 

* Appendix 4 presents an inrerpersonal diagnostic report written for the psychiatric 
clinic. This report serves as a clinical illustration of the theory and methodology pre- 
sented in this chapter. The execution of multilevel interpersonal diagnosis is facili- 
tated considerably by the use of a printed booklet in which the tabular and diagram- 
matic steps involved in diagnosis are organized. This "Record Booklet for Inter- 
personal Diagnosis of Personality" also includes a simplified procedure for calculating 
interlevel discrepancies (variabihty indices), for plotting them on a profile sheet, and 
for preparing verbal summaries of the indices. A copy of this booklet is presented in 
Appendix 4. 



legal situation, to take an extreme example, the diagnostician is usu- 
ally called upon to determine whether the patient is grossly psychotic 
or not. The judicial authorities are generally not the least bit interested 
in the fine shadings of ego organization or the complexities of the 
oedipal situation. A single "yes" or "no" as to the sanity of the sub- 
ject is generally sufficient. 

Many psychiatric centers employ broad categories which are just 
as gross and dichotomous. Some admitting wards, observation cen- 
ters, etc., have the sole mission of deciding where the patient will be 
routed. If the patient is markedly psychotic, he goes to a psychiatric 
hospital; if he is neurotic, he is sent back to the community. For such 
purposes, these agencies do not require subtle differential, diagnostic 

In most psychiatric hospitals the diagnostic decisions are somewhat 
more complex. Differential diagnosis relates to differential treatment 
plans. One type of patient may be assigned to electric shock therapy, 
another to insulin treatment. Therapies of this sort are aimed at 
symptom removal and not character reorganization. The diagnostic 
and prognostic terms, therefore, tend to focus on the descriptive or 
symptomatic aspects of behavior. An interpersonal or characterologi- 
cal diagnostic system is, in this case, not the most relevant tool. In de- 
ciding between two types of physical therapies the nature of the pa- 
tient's interpersonal reflexes, the type of repressed motivation or the 
expected transference are not the central criteria. 

Descriptive diagnosis as presented in psychiatric textbooks is gen- 
erally considered to be crude, unreliable, and nontheoretical (1, 2, 3, 
5). The main reason for the slow progress in psychiatric nosology 
parallels, perhaps, the general crudeness of most psychotherapeutic 
techniques. Medical diagnosis is, by comparison, extremely specific 
and definitive because of the differentiated maturity of medical knowl- 
edge. There are hundreds of detailed medical diagnoses all pointing 
to specific medical treatment plans. By contrast, psychological theory 
and psychiatric practice is most limited. This situation seems, how- 
ever, to be improving. 

R. E. Harris, for example, points out: "With progress in psycho- 
therapy, diagnosis is becoming more and more a matter of assaying 
'therapeutically relevant' variables, i.e., those which are related to 
what happens in therapeutic interviews — the resistances, the defenses, 
the strength of the ego, the amount and quality of the anxiety and its 
sources, the quality of reality testing, etc. These variables are defined 
and understood both as they are inferred from the historical recon- 
struction of the life history and, more importantly, as they appear in 
the interaction between therapist and patient." (4, pp. 27-28) 



The modem American psychiatric clinic is taking on an increasing 
number of prognostic decisions. A survey of the intake procedures 
of the Kaiser Foundation Psychiatric Clinic revealed that there are 
over twenty ways of disposing of a case. Most of these decisions are 
based on characterological or interpersonal factors: the amount of ego- 
alien anxiety, the interpersonal techniques for handling anxiety, the 
kind and amount of motivation for personality exploration and change. 
Certain "pure" hysterics and psychosomatic patients whose bland, ego- 
syntonic, hypernormal adjustments mobilize against psychotherapy 
might be sent back to the referring physician with recommendations 
for supportive medical handling. Other hysteric or psychosomatic 
patients whose personality patterns reveal underlying feelings of de- 
pression, anxiety, or deprivation might be referred to specific, care- 
fully delimited therapeutic relationships. Patients are assigned to sev- 
eral kinds of individual psychotherapy depending on the nature of 
the personality picture. Three types of group therapy are available 
for certain kinds of patients. 

In a psychiatric clinic of this sort interpersonal predictions which 
define the amount of anxiety and the interpersonal patterns to which 
it is attached are most useful. For patients beginning the long road of 
orthodox psychoanalysis, interpersonal diagnosis is less relevant. 
Transference factors are provoked and dealt with during the lengthy 
process, and preanalytic predictions might not necessarily save any 
time or energy. In the clinic, however, where flexibility of treatment 
program is emphasized, predictions as to expected behavior, expected 
resistances, and cues concerning the nature of repressed motives are 
at a premium. Interpersonal variables which measure the social re- 
activity of the patient, overt and covert, current and future, appear 
to have the highest functional "cash value." 

From the standpoint of practical application and research objectiv- 
ity, interpersonal diagnostic patterns seem superior to Kraepelinian or 
psychiatric diagnostic terms. This is not to say that standard psychi- 
atric nomenclature should be or could be abandoned. There are, in- 
deed, several factors which argue against the offhand rejection of 
psychiatric language. In a following section we shall seek to discover 
relationships and communalities between interpersonal and psychiatric 
diagnoses. Such relationships, if they exist, would greatly broaden the 
functional and theoretical power of both diagnostic systems. 

Any personality diagnosis, thus, serves several ends. It provides a 
classification most useful for administrative, legal, predictive, and 
research purposes. A more detailed interpersonal diagnosis serves the 
added function of predicting the kind and sequence of security opera- 
tions to be expected from the patient. 


In the following pages we shall present a diagnostic system com- 
prising 65,536 categories. These are based on the permutations of 
interpersonal measurements "adjustive and maladjustive" at three lev- 
els (four layers) of behavior. This is an automatic diagnostic pro- 
cedure. It is based on the logic, theory, and methodology which have 
been discussed in the preceding chapters. There is no clinical judg- 
ment or psychological intuition involved in making an interpersonal 
diagnosis. The personality data are collected and fed into the nota- 
tional apparatus, and the diagnosis automatically rolls out. In the 
Kaiser Foundation Clinic the tests of Level I and II are scored by 
clerical workers; the themes of Level III are rated by nonprofes- 
sional technicians. The resulting indices are plotted on standardized 
graphs, and a diagnosis involving three levels of interpersonal be- 
havior, as well as a diagnosis of variability, is obtained. Not one minute 
of professional time is required for these systematic diagnoses. (An 
independent clinical diagnosis is, of course, prepared by the psychiatric 
intake worker, and the two assessments of the patient are fitted to- 
gether in the clinical evaluation conference.) 

Systematic diagnosis based on multilevel test batteries is much more 
precise and detailed than clinical diagnosis. It is quite difficult to ob- 
tain reliable diagnoses when clinicians use only their own percep- 
tions and observations. Studies of the reliability of psychiatric judg- 
ments are notoriously low even when the simplest, broadest fourfold 
categories are employed (1, 2, 3, 5). If clinicians were asked to select 
a diagnosis from a list of 65,536 categories, an endlessly long period of 
intensified training would be necessary to effect any kind of reliability. 

The systematic diagnoses we are about to consider require, how- 
ever, no intuitive decision. The categorization system works some- 
thing like a table of logarithms or square roots. Much time and some 
creativity have gone into the development of these mathematical tables 
— but to use them is a routine, clerical, and reliable procedure. What- 
ever clinical experience and theoretical competence was available 
has been built into the classification schema. The application of the 
system is a rote process. The professional worker then takes the re- 
sults of the systematic operations and fits them to his clinical knowl- 
edge of the case. 

Three Systematic Methods for Summarizing Personality 

The meaning and function of diagnosis will be made clearer if we 
compare it with the other methods of summarizing personality data. 
The Kaiser project has employed three different kinds of assessments 
of human personality — each with its own purpose and significance and 
each possessing certain limitations and advantages. These are: (1) 


the diagnostic code-formula of personality, (2) the diagnostic profile 
of personality, and (3) the diagnostic description of personality. 

The diagnostic code is a succinct, systematic multilevel label or 
coded formula which is taken from a finite standardized list of mutu- 
ally exclusive terms. These terms should be operationally defined. 
The diagnostic classification should, therefore, be highly reliable. Its 
primary purpose is to summarize the essence of the multilevel pattern 
of personality, to make predictions with known probability about cer- 
tain crucially important aspects of behavior, and to prognosticate the 
success of specific clinical, therapeutic techniques. 

The multilevel diagnostic code can be calculated by well-trained 
clerical workers. The predictions can also be produced by clerical 
procedures since they involve the looking up of probability indices 
based on group statistics. A highly skilled clerical worker with no 
knowledge of psychological theory could hypothetically assemble the 
raw data, feed them into the tables and indices, and arrive at the auto- 
matic multilevel code diagnosis. This worker could then enter a set 
of correlation matrices and make probability statements about the 
patient. Statements of the following nature could be made: 'This pa- 
tient is an overtly autocratic personality; less than 20 per cent of pa- 
tients with this Level I diagnosis enter individual psychotherapy; of 
those patients who do enter individual psychotherapy, more than 65 
per cent quit within six weeks; of those who enter group therapy, 
more than 40 per cent quit within six weeks; etc." 

The advantages of the diagnostic code include brevity, reliability, 
finite listing of possibilities, and the fact that it does not require the 
expenditure of professional energy. The disadvantages are numerous: 
it has no explanatory value; it predicts only for a group (i.e., in 
probability figures) and takes no account of the unique complexity of 
each individual. 

The diagnostic profile of personality presents a diagrammatic and 
numerical summary of the patient's behavior at each level, and of his 
variability indices. It summarizes his behavior in terms of the inter- 
personal measurements, and in terms of the operationally defined 
indices of personality organization. The procedures on which the 
profile is based have a known reliability. 

The diagnostic profiles can be prepared by well-trained technicians, 
i.e., semiprofessional workers with a rudimentary knowledge of per- 
sonality theory. These technicians, in the Kaiser project, administer 
the testing batteries. They supervise the scoring of the Level I and 
II tests. They work in teams to score the fantasy material in terms of 
the interpersonal variables. They perform the necessary arithmetic 
procedures — calculating the horizontal and vertical indices, convert- 


ing them to standard scores. They plot the interpersonal scores on 
the diagnostic grids. They measure the discrepancies among these 
scores and thus determine the amount of the variability indices. They 
plot the variability profile. 

With these two diagnostic patterns — the variability and interper- 
sonal profiles — it is possible to classify patients in a most detailed man- 
ner. Eight interpersonal indices (at four levels) and 14-18 variability 
indices are available for automatic interpretation. These technicians 
are capable of translating this matrix of scores into statements which 
summarize the personality. They can report the behavior at Levels 
I, II, III, and V. They can indicate the amount of each variability 
mdex. A considerable mass of finely graduated information can be 
routinely reported for clinical or research purposes. In addition to 
these detailed classificatory statements, the personahty profiles al- 
low for a battery of individual predictions. Each of the eight inter- 
personal and the many variability indices have a set of empirical facts 
related to them. The Level II versus V discrepancy (self-description 
versus ego ideal) if high defines low self-acceptance or high motiva- 
tion for treatment. This is correlated highly with entering and remain- 
ing in psychotherapy. This variability index, thus, leads to a specific 
clinical prediction. The other indices have similar prognostic applica- 

The personality profile serves, in this way, as a precise, detailed 
classificatory system, and as a source for numerous specific clinical pre- 
dictions. It has, however, limited explanatory value. It is also re- 
stricted because of its routine objective nature. The personality pro- 
files are rehable — they do not involve speculation or intuition. The 
variability indices are ground out automatically, but they fail to inte- 
grate the complex network of scores. They do not in any way lessen 
the necessity for clinical, professional interpretation. They rather 
serve as a highly articulated assistance to the clinician. 

The personality diagnostic report.^ In the Kaiser Foundation Clinic 
the professional diagnostician enters the picture after the personality 
profiles have been plotted. All the testing, scoring, tabulating, and 
statistical predictive procedures are accomplished by technical work- 
ers. The task of the professional clinicians is to weave the multi- 
dimensional pattern of scores and probabiUty statements into a unique, 
meaningful summary which fits a particular patient. This is ac- 
complished by means of the personality diagnostic report. 

The main task of the diagnostic classifications is to categorize re- 
liably. The profiles provide a long list of standardized probability 

* A sample diagnostic report written about an illustrative patient is contained in 

Appendix 4. 


statements. The diagnostic report has the function of explaining the 
personality. The clinician studies all aspects of the interpersonal pro- 
file and relates them to the case. 

The intake worker's notes on family history are compared with 
the patient's view of mother, father, and spouse — and with his fantasy 
images. The indices of motivation and prognosis are compared with 
the patient's symptomatic presentation. All the available clinical data 
are reviewed in light of the personality profiles and indices. In this 
way, the experience of the clinician is brought to bear on the evalua- 
tion process. The thousandfold, multifaceted pattern of the patient's 
situation can be assembled in the mind of the professional worker. 

The diagnostic evaluation and the predictions are, of course, in- 
creased in efficiency as the clinical material qualifies and amplifies 
the conclusions of the more routinized profiles. The diagnostic report 
allows room for the creativity and insight which no systematic schema 
can duplicate. The profiles are, of course, analogous to the laboratory 
and radiological indices provided to the medical diagnostician. The 
final diagnosis and prescription is based on the multiple correlation 
procedure of great complexity which takes place in the mind of the 
professional worker. 

The diagnostic report deals with such a vast array of cues — clini- 
cal and systematic — that it is highly individualized. It is less reliable 
than the classification and profile ratings. It has a margin of unrelia- 
bility which must be hazarded because of the complexity of the sub- 
ject matter. 

From the research standpoint, the personality report possesses a 
great value. The correlations and results by which we test yesterday's 
hypotheses are provided by the objective indices of the interpersonal 
and variability profiles. In our scientific validation procedures we do 
not rely on the intuitive personality report for proof. The qualitative 
clinical report has, however, an inestimable research value because it 
produces the hypotheses of tomorrow. Scientific progress in personal- 
ity psychology works upward from the clinical, creative speculations 
which are first expressed in the personality reports. The profiles, it will 
be noted, serve admirably to give objective tests to hypotheses and to 
yield probability predictions. Their very objectivity, however, guar- 
antees that they will never generate a new idea. 

The future of personality research lies in the front lines of the 
functional situation. The neat predictions from the research office get 
dented and pushed around by the rough pressure of human individual- 
ity. We know very well that the indices do not take into account the 
familial, occupational, cultural, or educational history of the patients 
for whom they attempt to predict. At exactly the spots where the 



z d 
< o 

(So <>J 

15 t 5J a 

bo c^ 
■> c - ... 
n u V 

w "^ C S 

O 3 
"C O 

S 5 .- 
>< i-i n 

S 1^ 

5 t! = 2 a. 
U u C a. " 

c o o " 2 «J 
T3 -D c </) E Cc 
S S ^:2 E 2 
OC^'-B So « CL 

.2^ p ^-a i § 

•£ c o u u, .5 

b o 8 =3 I g 2 


2 ^u 
i E 3 

< S y 


« O O 4J i2 C 2 

n 60 ^ u 2 y 

u 2 c 2 Ji ul 3 

Q- rt o <u S-y S 

c w — 
W J^ S il 

o Ch' o -a ' 

-o u E " : 

S a-O b' 

c c ."H x: 


w <0 C -o 

i'E «^:2 

> rt S i; n 

X p-2 

is E 

S 2 I 



predictive indices fail, the clinician is present to observe and to inte- 
grate. If the predictions hit an accuracy of 90 per cent they still fail on 
ten patients in every hundred. Each ten failures may be related to new 
clinical variables which may later be added to the system, recast in 
operational language, and tested for validity. Inaccurate predictions 
are not embarrassments but sources of new hypotheses. 

All three elements of diagnosis — the classification, the profiles, and 
the report — are necessary for optimal evaluation of personality. Each 
has its limitations — each makes its unique contribution to the clinical 
and research aspects of the science. Many problems and frictions in 
psychology might be lessened if the nature and function of these three 
elements is kept straight. Table 6 presents a summary of these three 
types of diagnostic procedures which are now employed in the Kaiser 
Foundation project. 

Functional Diagnosis 

The evaluation of personality which we are describing in this book 
is called functional diagnosis. The aim of our measurements is to un- 
derstand the patient-in-his-relationship-to-the-clinic and to make pre- 
dictions about the patient-in-relationship-to-his-future-therapist. The 
focus of our diagnostic observations is the interpersonal behavior in 
the context in which we (the clinical staff) have commerce with him. 
The results of a psychological test battery can be studied from many 
vantage points. It is possible for psychologists to predict competence 
in combat flying, academic success in college, occupational fitness, etc. 
These predictions may, in certain situations (in the air force, on the 
campus, in the factory) , be quite functional and relevant. They would 
not be much help to the psychiatrist attempting to decide the best plan 
of psychotherapy. 

Psychologists often use clinical tests to estimate the patient's 
creativity, constriction, impulsivity, etc. These findings have some 
value in the understanding of personality, but they would be of little 
use to the clinician who wants to know: "Why is this patient coming 
to the clinic? How much and what kind of motivation is present? 
How will he react to different types of treatment? " 

Other psychologists are able to outline with impressive sophistica- 
tion the probable genetic history of the patient, the early traumatic 
events, and the finely detailed nuances of the patient's sexual adjust- 
ment. For patients who enter psychoanalysis or long-term treatment 
these predictions undoubtedly point ahead to materials which will 
emerge in the associative content. The Kaiser Foundation project has 
not attempted to measure these areas of personality — partly because 
they are beyond the scope of our technical capacity, and partly be- 


cause we believe them to be less functional in regard to the crucial 
aspects of pretherapy planning and of therapeutic interaction. 

The first aim of functional diagnosis is to summarize before treat- 
ment the aspects of the personality which have a bearing on the choice 
of treatment. What is the motivation of the patient in coming to the 
clinic? Does he come with self-depreciation, ready to unburden his 
innermost thoughts and expecting some kind of mystical cure to fol- 
low his confidences? Disappointment and bitter reproach may be the 
easily predicted outcome if this motivation is not perceived and 
planned for. Does he come under pressure from someone else (e.g., a 
physician), defensively mobilized against any self-examination? A 
stubborn power struggle and angry departure may be predicted if this 
motivation is not recognized and responded to. 

Clinical diagnosis concerns ego factors which influence the choice 
of treatment. How much anxiety is manifested? What are the se- 
curity operations by which the patient handles anxiety? What is the 
interpersonal pressure put by the patient on the chnic? 

The first aim of functional diagnosis is, then, to assess motivation for 
treatment. The second aim is prognosis of treatment — to summarize 
the kind of behavior which will appear in future therapy. How fast 
or slow will be the course of therapy? Many patients who are well 
motivated for change (thus satisfying the first criterion of functional 
diagnosis) also manifest chronic, deeply rooted security operations 
which are most resistant to change, or underlying psychotic distrust 
which had best be left unexplored. 

Another aspect of prognosis concerns the nature of the intrapsychic 
conflicts. In many cases it is possible to point to private or "precon- 
scious" motives which will probably afl^ect the later treatment rela- 
tionship. A different transference relationship and prognosis are indi- 
cated depending on whether the underlying themes involve distrustful, 
passive resistance, or independent autonomy, or nurturant tenderness. 

In making our prediction about the first (or motivational) aspect of 
functional diagnosis we are mainly interested in "ego" factors. In 
making the second prediction we concentrate on the deeper, "pre- 
conscious" aspects of personality and their relationship to the more 
overt or public factors. We use the total interpersonal profile to map 
out areas of anxiety, the security operations by which it is handled, and 
the transference phenomenon which they will tend to elicit during 

Interpersonal Typology 

To accomplish functional diagnosis the total personality pattern 
(both the interpersonal and variability profiles) is employed. In- 


volved in this total matrix are thousands of individual measurements 
which are summarized in terms of sixteen interpersonal diagnostic 
circles and an extensive assortment of variability indices. The com- 
plexity of this system of diagnosis is such that it is almost impossible 
that any two patients would ever show exactly the same kind of multi- 
level profile. 

For this reason a typological system is needed. jMethods have been 
developed for summary classification of interpersonal and variability 
types. In this chapter we shall consider the interpersonal typology.^ 

Two Kinds of Interpersonal Typologies. We recall that inter- 
personal behavior has been rated at five levels. Two of these — Level I 
communication and Level II self-description — are considered to refer 
to the overt picture which the patient expresses in the clinical situa- 
tion. These are designated presenting operations. A technique has 
been worked out for typing or categorizing the more overt, conscious 
aspects of the patient's personality. We shall speak, for example, of 
the managerial type or the over conventional personality. 

The categories which serve to summarize these behavioral opera- 
tions do not directly apply to the underlying motives of the patient. 
The "preconscious" themes of Level III tend to require a different 
typological language. In an earlier chapter we have noted that sub- 
jects in their fantasy themes express motives which are more intense 
and extreme than those which they manifest in their overt behavior or 
conscious self-descriptions. For this reason it is necessary to develop 
a typology for summarizing the underlying operations of the patient. 
The presenting operations (of Levels I and II) are, therefore, sum- 
marized in terms of behavioral types, and the underlying operations 
are categorized in terms of thematic motives. The specific diagnostic 
classes of these two aspects of personality will now be presented. 

Interpersonal Diagnosis of Presenting Operations. The first 
goal of interpersonal diagnosis is to summarize the overt behavioral 
impact of the patient upon the clinic. The task here is to determine: 
(1) the kind of security operations, (2) the adaptive or maladaptive 
pattern of conflict. This is accomplished as follows. First, the Level 
II Self and the Level I symptomatic scores are converted into hori- 
zontal and vertical indices and plotted on the diagnostic grid.'' 

These summary placements, it will be noted, reflect two aspects of 
the Level I and II behaviors. The sector of the circle tells us what 
interpersonal operations are involved, and the distance from the center 

^The variability dimension of personality is discussed briefly in Chapter 13. 

" The methodology for converting conscious self-descriptions and "symptomatic 
communications" into horizontal and vertical indices and for plotting these indices on 
the diagnostic grid has been explained in earlier chapters. 



tells us how extreme or intense they are. Figure 26 presents an illus- 
trative diagraming of the Level I-M and II-C scores for an overcon- 
ventional patient. Both the Level I-M and II-C scores fall in the M 
sector of the circle, and both fall in the outer ring of the circle, indi- 
cating that they are more than one sigma above the mean in the direc- 








\s^^ / \ 




/ \, \ 

/ \^ 


° ,,, 1 ,,,,?,,, , 





\ /. 






Figure 26. The Summary Placements of the Level I-M and II-C Scores for an 
Overconventional Patient. Key: The center of the grid js determined by the intersec- 
tion of the means of the honzontal and vertical distributions. The grid is calibrated in 
standard score units. 

tion of conventionality and blandness. Two typological procedures 
allow us to classify this pattern. The first step is to summarize the 
sixteen-variable matrix into eight diagnostic categories. This is ac- 
complished by combining adjacent variables and assigning a descrip- 
tive term which reflects the interpersonal meaning. Pairing the sixteen 
variables yields eight sectors of the diagnostic circle, which are called 
octants. Thus we combine A (forceful dominance) with ? (respected 



success) into a power octant, and we combine B (self-confident inde- 
pendence) with C (competitive self-seeking) into a narcissistic octant. 
We also express in our diagnostic summaries the adjustive or mal- 
adjustive aspects of the presenting operations. If the subject's behavior 
falls in the outer ring of the circle (one sigma above the mean), an 
intense or maladaptive degree of this interpersonal behavior is indi- 
cated If his security operations fall within one sigma of the center of 
the circle, a moderate, adaptive degree is indicated. A statistical tech- 
nique for diagnosing normality-abnormality or adaption-maladaption 
in terms of degree is thus available. Figure 27 presents a schematic 









personality/^ ^v^ PERSONALITY 






"^v / SELF- 












Figure 27. The Diagnosis of Interpersonal Behavior at Levek I and II. 

illustration of this method of interpersonal diagnosis and illustrates the 
fact that there are eight interpersonal diagnostic categories by which 
we summarize the overt behavioral presentation of the patient and that 


there is an adaptive and a maladaptive degree to each type. These eight 
adjustive and maladjustive categories are formally listed in Table 7. 

The Adaptive and Maladaptive Interpersonal Diagnostic Types 

The Adfiistive Inter- 
personal Types {one 
sigma or less from 
Interpersonal the mean) and the Nu- 

Variable merical Code Used to 

Code Designate Them. 

AP 1 Managerial personality 

BC 2 Competitive personality 

DE 3 Aggressive personality 

FG 4 Rebellious personality 

HI 5 Self-effacing personality 

JK 6 Docile personality 

LM 7 Cooperative personality 

NO 8 Responsible personality 

The Maladjustive Inter- 
personal Types {one sigma 
above the mean) and the 
Numerical Codes Used to 
Designate Them. 

Autocratic personality 
Narcissistic personality 
Sadistic personality 
Distrustful personality 
Masochistic personality 
Dependent personality 
Overconventional personality 
Hypernormal personality 

It v^^ill be noted that for each adaptive and maladaptive type there 
is a verbal descriptive category (e.g., conventional; overconven- 
tional) and a numerical code. The numerical designation for the con- 
ventional type is 7, and for the overconventional type it is 7. The 
numerical index is a most convenient w^ay of summarizing behavior at 
any level because the digits can be combined into diagnostic formulas. 
An italic diagnostic digit always refers to the moderate amount of the 
interpersonal trait. A roman-face diagnostic digit always refers to the 
intense or extreme amount of the behavior. This system of numerical 
code diagnosis will be discussed in later sections of this chapter. 

These diagnostic terms are employed to summarize security opera- 
tions at Levels I and IL The diagnosis is accomplished automatically 
by locating the indices for Level I in the correct octant sector and se- 
lecting the interpersonal term which reflects this particular behavior. 
This yields the Level I diagnosis. The same procedure is followed for 
Level II — the indices determine the proper sector, and the appropriate 
term is thus determined. 

A printed booklet for deriving a multilevel interpersonal diagnosis 
is presented in Appendix 4. 

Figure 28 presents an illustration of Level I and Level II diagnosis 
for two sample patients. 

The Level I-M scores for patient "X" fall within one sigma of the 
mean of the NO octant. Referring to Table 7 we see that this defines 
an adaptive degree of responsible behavior. Patient "X" is therefore 
diagnosed at Level \-\{ as a Responsible Personality (numerical code 


diagnosis = 8). The Level II-C score for this patient is in the same 
octant but beyond the normal range. This patient is self-diagnosed as 
a Hypernormal Personality (numerical code diagnosis =8). For rou- 
tine diagnostic research categorization of what we have called "pre- 
senting operations," the Level I and II diagnostic terms can be com- 
bined in a hyphenated designation, which for patient "X" would be 
Responsible-Hypernormal Personality. A more convenient summary 
method is to combine the numerical codes, which for patient "X" 
would be 8S. The first term (verbal or numerical) in any diagnostic 
formulation always refers to Level I-M symptomatic behavior, and 
the second term to Level II-C self-description. 

The Level I-M of the other illustrative patient "Y" falls in the outer 
ring of the HI sector. This (by reference to Table 7) is seen to define 
a Masochistic Personality at Level I-M (numerical code diagnosis 
= 5). The Level II-C scores fall in the extreme end of the JK octant. 
The Level II-C diagnosis is: Dependent Personality (numerical code 
diagnosis = 6). The combined diagnosis of presenting operations is 
Masochistic-Dependent personality, or more simply a ''56" personality 

In practice, the single level diagnosis is rarely employed. The focus 
of the functional diagnosis generally includes the combined presenting 
operations of Level I-M plus Level II-C. There are sixteen possible 
categories for diagnosis at Level I-M — eight adjustive and eight mal- 
adjustive types. The same number of categories are available for sum- 
marizing Level II-C behavior when we turn to the combined diag- 
nosis involved in the presenting operations. There are, therefore, 256 
categories for interpersonal diagnosis of presenting, or facade 

To present a diagnostic system involving 256 types would seem to 
be an audacious gesture, placing an impossible task on the diagnostician 
who attempts to use this system. It has already been pointed out that 
diagnosis in the interpersonal system is a most routine and unde- 
manding procedure. What we ask of a diagnostic label is an objective 
summary categorization for administrative or research classificatory 
purposes. We do not ask our diagnosis to take the place of a personal- 
ity profile or a personality description. It is, instead, an automatic and 
rehable classification. In practice, the interpersonal diagnosis is rou- 
tinely determined by clerical procedures. The patient's scores on 
Levels I and II are placed on the diagnostic grid, and the appropriate 
diagnostic terms or two-digit diagnostic codes are automatically de- 
termined. The theory and methodology of the interpersonal system 
is complex, but the employment of the system for the purpose of diag- 
nostic classification is simple. 



Figure 28. Illustration of Interpersonal Diagnosis at Levels I-M and II-C for Two 
Patients, "X" and "Y" {see facing page) . 

Interpersonal Diagnosis of "Underlying Operations." The sec- 
ond goal of functional diagnosis is, we recall, to summarize the the- 
matic "preconscious" motifs, to relate them to the overt presenting op- 
erations, and to employ these data to make predictions about the fu- 
ture course of treatment. We shall now consider a method for sum- 
marizing the "preconscious" themes of Level III. 

The theory and measurement methods for Level III fantasy expres- 
sions were presented in Chapter 9. It was emphasized that there are 
two layers of "preconscious" data — the themes assigned to fantasy 
heroes and those assigned to fantasy "others." These sublevels involve 
different psychological functions. They have different lawful rela- 
tionships with the other aspects of personality structure. These two 



















^(HiJ ■ 



Figure 28 (cont.) 

sublevels provide two diagnostic types of underlying operations. 

In Chapter 9 we presented a method of measuring the interpersonal 
themes from fantasy data and for converting the resulting scores into 
indices which locate the subject on the diagnostic grids for Level III-T 
Hero and Level III-T Other. There remains the task of formally di- 
viding the Level III-T grids into summary diagnostic sectors. This is 
accomplished in the same manner as for Levels I-M and II-C. The 
sixteen-variable matrix is combined into the same eight sectors and the 
appropriate descriptive terms or numerical codes assigned. The Level 
III diagnostic circle is presented in Figure 29. 

In comparing the diagnostic grid for Level III with that used for 
Levels I and II, two differences will be observed. The first concerns 
terminology — Level III deals with underlying private motives and 
not behavioral manifestations. The diagnostic terms tend, therefore, 
to reflect general motivational purposes rather than social role be- 

2 24 











Figure 29. The Diagnosis of Interpersonal Behavior at Level III Hero and Other. 

havior. A second difference concerns the degree of intensity. At 
Levels I and II we distinguish between adaptive behavior and mal- 
adaptive extremes. At Level III no such distinction is maintained in 
setting up verbal descriptive categories. There are two reasons: Theo- 
retically it is questionable that "preconscious" imagery can be con- 
ceptualized as being adaptive or maladaptive. In one sense it appears 
that this differentiation violates the notion of the equilibrium or 
"safety valve" function of fantasy. In some cases the most violent 
and antisocial autism may serve a healthy balancing function. The sec- 
ond reason for not making the adaptive-maladaptive distinction at 
Level III tends to make the first argument academic. At this point our 
measures of Level III fantasy are so crude and preliminary that the 
fine distinctions of normal versus abnormal autisms have been difficult 
to study. The distinction between italicized numerical digits (for 


moderate behavior) and roman-face digits (for intense behavior) is 
maintained at Level III, but it simply reflects the amount of the inter- 
personal emotion. 

There are, then, eight verbal summary categories and sixteen nu- 
merical categories for diagnosing behavior at Level III Hero and Other. 
We have been continually reminded by one of the basic mottoes of 
this book that no level has meaning by itself, and each level must be 
interpreted in light of the other levels. Some theories of personality 
do tend to diagnose on the basis of underlying motives. The Kaiser 
Foundation system, on the contrary, does not follow this procedure. 
We may tend to overemphasize the presenting operations (i.e., diag- 
nosing from Levels I and II), but complete diagnosis includes the un- 
derlying thematic behavior. 

Multilevel Interpersonal Diagnosis. We are now ready to 
proceed to the complete interpersonal diagnosis of the two sample 
cases whose behavior we have been considering in this chapter. Inter- 
personal diagnosis is accomplished by combining the sumiTmry descrip- 
tive terms for Self behavior at Levels I and II, and the Self and Other 
behavior at Level III. The significant omissions of Level IV, and the 
behaviors of ''others'' at all levels (except III), and the value-themes of 
Level V are not included in the diagnostic categorization. It must also 
be kept in mind that we are dealing, in this chapter, with the inter- 
personal aspects of diagnosis. The variability diagnosis, which com- 
prises the other half of functional diagnosis, will be treated in Chap- 
ter 13. 

Figure 30 presents the Self scores of two patients, "X" and "Y," at 
the three levels (I-M, II-C, and III-T) which comprise interpersonal 
diagnosis. The methodology for plotting the summary scores on the 
diagnostic grid has already been described. The diagnostic classifica- 
tion is automatically obtained from Table 7 (for the Level I-M and 
II-C terms) and Figure 30 (for the Level III terms). The Level I-M 
score for patient "X" falls in the adaptive ring of the NO octant (nu- 
merical code = S), his Level II-C score in the maladaptive range of the 
same octant (numerical code = 8), the Level III-T Hero and Other 
scores in the DE octant (both coded 3). His interpersonal diagnosis 
is: responsible-hypernormal personality nvith underlying hostile feel- 
ings. The four digit diagnostic code is 5*833. 

The Level I-M and II-C scores for Patient "Y" fall in the mal- 
adaptive ring of sectors FG and JK respectively (coded 46). The 
"preconscious" Hero scores are in the outer ring of the BC octant 
(coded 2) and the "preconscious" Other scores are in the inner ring 
of the HI octant (coded J). Referring to Table 7 (for the Level I 






^/§/ m-T 












\^ ^ T 

r "" ■/ 


/ \ 






|. M?M, 

, , , '|° , M 1 , , , ,' 






[\ /^\ 


Figure 30. Illustration of Interpersonal Diagnosis at Levels I-M, II-C and III-T 
for Two Patients ("X" and "Y"). Key: For Levels I-M and II-C the circular dotted 
line indicates the distance of one standard deviation from the mean. (The mean for 
both the honzontal and vertical distributions in the center of the circle.) At Levels 
1-M and II-C scores falling vi'ithin the inner circle are considered adaptive and those 
in the outer circle maladaptive. At Level III-T the verbal distinction between adaptive- 
maladaptive is not maintained. Figure 29 presents the interpersonal diagnostic cate- 
gories for Level III. 

and II terms) and Figure 29 (for the Level III terms), we see that his 
interpersonal diagnosis is: distrustful-dependent personality with un- 
derlying narcissism and deeper feelings of masochism. The diagnostic 
code for this four-layer pattern is 5625. 

The verbal diagnostic formula comprises a four-part sequence. The 
Level I diagnosis comes first, and it is always paired with the Level II 
diagnosis. The hyphenated description summarizes the presenting op- 
erations of the patient. The Level III Hero and Other diagnoses fol- 















\v / \ 


^ '\ 


, . , 2,0, , ,' 


1 , , , ,'m . , , 

, , , ,•<» ,« 




\ /n-{ 








Figure 30 (cont.) 

low. The standard grammatical structure of an interpersonal diag- 
nosis can now be considered: 

" - Personality with Underlying , 

AND Deeper Feelings of 

The first two elements, 

personality," are 
The third element 
," is obtained 

determined by the Level I and II summary scores 

in the diagnostic formula, "with underlying 

from the Level III Hero summary scores. The fourth element comes 
from the Level III Other indices. To complete the functional diag- 
nosis we add the variability diagnosis. 

The Numerical Diagnostic Code. The numerical diagnostic 
summary is a four-digit formula which sequentially presents the Level 
I, II, III Hero, and III Other scores. 


This four-layer system of interpersonal diagnosis provides an ob- 
jective, reliable, and standardized classification. The diagnostic for- 
mula summarizes the four aspects of the patient's security operations 
which are most central to the decisions which clinicians make about 
patients: How does he act? How does he see himself? What are his 
underlying interpersonal potentials? 

A detailed and complex diagnosis is involved in the four-layer 
formula. It was pointed out that there are 256 types of presenting op- 
erations. Each of these 256 diagnostic types can be characterized by 
any one of eight Level III Hero patterns and by any one of eight Level 
III Other patterns. For example, the patient whose presenting opera- 
tions are diagnosed "responsible-hypernormal" can have underlying 
"preconscious" themes located in any of the eight Level III Hero 
sectors and in any one of the eight Level III Other sectors. There are 
64 combinations of Level III Hero and Other scores. There are, there- 
fore, 16,384 (256 X 64) verbal diagnostic formulas available for 
summarizing human security operations. When we consider the four- 
layer combination of numerical codes, a much larger set of possibili- 
ties exists. It will be recalled that the distinction between moderate 
and extreme behaviors is preserved in numerical diagnosis, italic num- 
bers referring to the former and roman-face numbers to the latter. 
There are, therefore, 65,536 (256X16X16) numerical formulas 
available for summarizing interpersonal behavior at four layers of 

Of the 16,384 verbal diagnostic categories, one quarter characterize 
the adaptive or adjusted personality (at the level of presenting opera- 
tions). One quarter of them involve maladaptive types. One half of 
them designate personalities who have inconsistent ratings of adjust- 
ment-maladjustment. Of this latter group it is obvious, by definition, 
that half of them are self-diagnosed as adaptive in the context of a 
symptomatic (Level I) diagnosis of maladjustment, while the other 
half of this mixed group are self-diagnosed as maladjusted in the 
context of a clinical symptomatic rating of adjustment. 

Each of these four broad categories has an obvious clinical and 
theoretical meaning. A diagnostic formula is, of course, a crude and 
rough estimate of the security operations of the individuals, and it in 
no way can be substituted for the more detailed pattern obtained from 
the interpersonal and variability profiles or from the personality report. 
With this qualification in mind, it can be seen that a specific multilevel 
diagnosis (of the sort we have just presented) provides a useful core of 
relevant information. In one standardized four-digit formula we ob- 
tain a summary of (1) the interpersonal behavior at three levels and 
(2) an estimate of kind and degree of adjustment-maladjustment. 


The Relationship of Interpersonal Diagnosis to 
Pyschiatric (Kraepelinian) Diagnosis 

In an earlier chapter it seemed pertinent to comment on the prob- 
lems created by the novelty of the interpersonal system. When the 
reader who has struggled to acquaint himself with sixteen variables, 
five levels, fourteen variability indices is now presented with 16,384 
verbal diagnostic types or 65,536 diagnostic codes, the proliferation 
of new terms and concepts may seem to be getting out of hand. At 
this point it may prove encouraging to point out that the interpersonal 
diagnostic system is not completely divorced from the traditional 
language of the clinic. In our diagnostic formulations we employ a 
behavioristic and interpersonal language to summarize the patient's per- 
sonality. We are often called upon to communicate our diagnoses to 
other clinicians who are not familiar with the systematic language of 
the interpersonal system. 

The interpersonal language — masochistic, autocratic, etc. — has 
the advantage of a narrow and parochial usage. The Kraepelinians' 
diagnostic language, by comparison, possesses an almost universal 
popularity of usage, but is often vague and unclear. Most every 
clinician tends to use labels such as schizoid or hysteric, but many of 
them mean quite different things by these terms. One clinician may 
conceive of the hysteric in terms of certain colorful symptoms; an- 
other may refer to the state of psychosexual development; another 
may denote a certain pattern of repressive defenses when he employs 
the term. 

However obvious its drawbacks, the Kraepelinian nosology has two 
irrefutable claims to survival — its widespread acceptance and its im- 
plicit connotations. A great deal of wisdom has accumulated in the 
folklore of psychiatry. By this we mean that a psychiatric label 
(e.g., hysteric) is a crude, disorganized synthesis of many variables 
of behavior. Some of these are inconsistent, some subjective to the 
user, some tautological, some unimportant, some valuable. Further, 
it seems safe to say that most psychiatric labels have some interpersonal 
factor loading. Schizoids show different interpersonal behavior from 
hysterics, or from phobics, or from obsessives. It follows, therefore, 
that there must be considerable overlap between the standard Krae- 
pelinian nosology and the interpersonal diagnostic system presented 
in this chapter. 

With this hypothesis in mind an investigation of the interpersonal 
factors in psychiatric terminology was undertaken by the Kaiser 
Foundation research project. If this hypothesis is true, then several 
advantages will accrue to both diagnostic systems. The objectivity. 


reliability, and systematic complexity of the interpersonal schema can 
be related to the standard Kraepelinian terminology. The latter might 
be partially defined in terms of the operational language of the for- 
mer. The pathological, maladjustive emphasis of the Kraepelinian sys- 
tem may be amplified by the adaptive dimensions of the interpersonal 

Testing the Relationship Between Interpersonal 
and Psychiatric Diagnosis 

The question now posed might be worded as follows: What is the 
relation between interpersonal and standard psychiatric diagnosis? The 
preceding sections of this chapter have presented a method for 
establishing interpersonal diagnosis at three levels (four sublevels) of 
personality. For comparison with Kraepelinian categories it seemed 
logical to employ the levels which define presenting operations, i.e., 
Levels I and II. For the exploratory investigations it was decided to 
use the Level II self-description as the interpersonal diagnostic cri- 
terion. The obtaining of an index of interpersonal diagnosis was, thus, 
a single straightforward task. The next problem was to find a 
measurement or rating of Kraepelinian diagnosis to compare with the 
interpersonal criterion. This was not as easy a procedure. The most 
obvious solution was to ask psychiatric clinicians to make diagnostic 
judgments of the same patients where Level II profiles were employed 
for interpersonal diagnosis. This proved to be unfeasible for two 
reasons. In the first place, psychiatric diagnostic judgments are no- 
toriously unreliable (1, 2, 3, 5). If one side of the comparison is an 
undependable measure, the extent of the true relationship between 
the variables is clouded. A second disadvantage of clinical diagnostic 
ratings concerns the subjective factors unique to each clinician. Ex- 
tended conversations with practicing clinicians revealed a wide varia- 
tion in individual preferences for use or avoidance of certain Krae- 
pelinian terms. Some psychiatrists expressed doubt as to their ability 
to employ certain diagnostic categories satisfactorily. Others believed 
that pure Kraepelinian types were rare, and that most patients show a 
mixture of reactions. Most all of them expressed a preference for dy- 
namic or psychoanalytic language. 

The search for a criterion measure of psychiatric diagnosis which 
would be reliable and standardized led us in the direction of psycho- 
metric estimates. The Minnesota Multiphasic Personality Inventory 
seemed to be a most satisfactory estimate of psychiatric diagnosis be- 
cause it is reliable and because there are widely accepted patterns of 
scores which are valid estimates of psychiatric diagnosis. 



The comparison procedure can be briefly summarized as follows. 
The MMPI diagnosis of 200 clinic patients was determined (by pooled 
ratings of three psychologists) . 

Six of the most common neurotic types or character disorders were 
employed as the diagnostic criterion. The Level II diagnoses of these 
same patients were obtained. The results are summarized in Table 8. 
They indicate that these six psychiatric diagnostic types are related to 
different interpersonal modes of behavior, 

Median Interpersonal Self-Description Score for Six MMPI Clinical Groups 



on Level 




II Inter- 

Summary of 





Level II 







F, Pd, Ma 



Aggressive, sadistic 


D, F, Sc, Pd 



Bitter, distrustful 


D, Pt 



Self-derogatory, passive 


D, Pt, Hy 



Docile, dependent 


Hy, K 



Bland, overconventional 


Hs, Hy 



Responsible, hyper- 
normal, generous 

Patients who employ aggressive, nonconventional modes of mal- 
adjustment tend to obtain the psychiatric diagnosis of psychopathic 
personality. Distrustful, passively-resistant modes of adjustment tend 
to be called schizoid; submissive, self-punishing patients tend to be 
called obsessives; docile, dependent patients tend to be labeled phobics; 
bland, naive, overconforming patients tend to be diagnosed hysterics; 
and responsible, hypernormal patients fit the psychosomatic pattern 
of the MMPI. 

Of the eight interpersonal modes of adjustment-maladjustment, six 
are related to psychiatric categories. Two interpersonal modes, how- 
ever, the autocratic-managerial and the competitive-narcissistic, seem 
to have no psychiatric equivalent. A new question arises: Why do 
two interpersonal modes fail to merit psychiatric diagnostic cate- 
gories? Why do managerial and competitive people fail to excite diag- 
nostic attention, and thus avoid formal psychiatric recognition? These 
categories are not unmentioned in the clinical literature. Exploitive, 
narcissistic, power-oriented techniques have been described by Fromm, 
Horney, and Sullivan. Prior to these culturally oriented writers little 
reference has been made to these cases. They have not obtained 
nosological popularity in any formal diagnostic system. 



We are led to speculate that these types have received little diag- 
nostic attention because they do not come for help. Perhaps they do 
not seek therapeutic assistance because the very essence of these mal- 
adjustments is a compulsive maintenance of autonomy, independence, 
and domination. These social techniques clearly preclude the role of 
a psychiatric patient. 

Our personality theories have generally been grounded in clinical 
practice. But there seems to be increasing evidence that major neu- 
rotic groups exist which are exposed to psychological testing diagnosis 
and therapy in disproportionately small numbers. 

Interpersonal Diagnosis. The data and speculations just pre- 
sented have encouraged the possibility of relating interpersonal and 
psychiatric diagnoses. It now seems feasible and profitable to define 
psychiatric diagnoses in terms of the interpersonal expressions of the 
patient. In this manner we preserve the values inherent in psychiatric 
diagnosis, its widespread acceptance, and its statistical, administrative, 
and theoretical advantages, which would be lost by a total rejection of 
classical terminology. For these reasons the Kaiser Foundation re- 
search project has retained the older clinical categories, combining 
them with, and defining them in terms of, interpersonal factors. 

Table 9 presents the eight modes of interpersonal adjustment and 
maladjustment and the suggested psychiatric categories to which they 
may be hnked. Under each interpersonal category we have listed not 
one trait, but a syndrome of behaviors which are most typical, and 
which often seem to go together. The category HI actually includes 
many normal responses — retiring modesty, thoughtful reserve, sensi- 
tive, deferent self-appraisal, etc. The same HI sector of the circle 
also includes an assortment of extreme, maladjustive reactions — pas- 
sive withdrawal, ruminative immobilization, submission, and self- 
punitive attitudes. A variety of psychiatric terms seems to be related 
to this generic interpersonal mode. Patients who fall in this area of the 
diagnostic circle are often clinically labeled masochistic, guilt-ridden, 
obsessive, or psychasthenic. This cluster of psychiatric terms is thus 
related to the interpersonal mode. Wherever syndromes of psychiatric 
categories or parallel diagnostic terms exist, they are included in 
Table 9. 

The relationships presented in Table 9 mark an important step in 
the segmental development of this book. They establish an important 
linkage between systematic interpersonal language and standard clini- 
cal terminology. They relate the standard categories of psychopath- 
ology to a continuum of maladjustive and (theoretically niore im- 
portant) adjustment types. 




Operational Redefinition of Psychiatric Categories in 
Terms of Interpersonal Operations 

ical Variable 
Code* Code 

Interpersonal Mode 
of Adjustment 

Interpersonal Type 
of Maladjustment 








Executive, forceful. Managing, auto- 
respected personality cratic, power-ori- 
ented personality 

Independent, com- 
petitive personality 

Blunt, frank, criti- 
cal, unconventional 

Realistic, skeptical 

Modest, sensitive 

Respectful, trust- 
ful personality 

Narcissistic, exploit- 
ive personality 

Aggressive, sadistic 

Passively resistant, 
bitter, distrustful 

Passive, submissive, 
self-punishing, maso- 
chistic personality 

Docile, dependent 

Standard Psychiatric 
Equivalent of Inter- 
personal Type of 

No psychiatric 
equivalent (Com- 
pulsive personality') 

No psychiatric 
equivalent (Counter- 
phobic' Manic') 

Psychopathic, sadis- 
tic personality 

Schizoid personality 

Masochistic, psychas- 
thenic, obsessive 

Neurasthenic, mixed 
neurosis, anxiety 
neurosis, anxiety 
hysteria, phobic 


Bland, conventional. Naive, sweet, over 
friendly, agreeable conventional per- 
personality sonality 

8 NO Popular, responsible Hypernormal, Psychosomatic 

personality hyperpopular, com- personality 

pulsively generous 

* The numerical codes for interpersonal diagnosis also designate adaptive or mal- 
adaptive intensity. Numbers in roman face refer to extreme maladjusted interpersonal 
behavior and italicized numbers denote an adjustive mode. 

We define "hysteric," "phobic," and other clinical diagnostic types 
in terms of the presenting operations of Levels I and II. It is immedi- 
ately possible to study the behavior of these diagnostic types at the 
other level of personality and in terms of the variability indices. How 
do hysterics (diagnosed at Level I-M) see their mothers, fathers, and 
spouses? What are the identification-disidentification indices for 
schizoids (diagnosed at Level I-M)? What are the Level V value- 
aspirations of obsessives, hysterics, etc.? What are the Level III 
fantasy patterns of psychosomatics? The eight clinical chapters of 
tliis book are entirely devoted to a consideration of these questions, to 


a summary of these new conceptions of diagnosis, and to their valida- 

Thus, in linking interpersonal terminology to psychiatric diagnoses 
we have facilitated a systematic investigation of many important clini- 
cal problems. It will be noted, however, that this research enterprise 
is based on what seems to be a rather shaky foundation — the equiva- 
lence of the Kraepehnian-type diagnosis to interpersonal patterns. An 
objection to this equation might point out that the relationship of 
MMPI diagnostic patterns to interpersonal types is far from being a 
convincing validation. Even though the MMPI is one of the most ac- 
cepted and popular diagnostic tests, and even though it is based on 
carefully diagnosed criterion groups, this one test cannot be considered 
a satisfactory criterion of psychiatric diagnosis. This objection is well 
taken. This is not a satisfactory criterion — but it still stands as the best 
criterion. The unhappy fact is that there is no possibility of getting a 
watertight estimate of Kraepelinian-type diagnoses. Like so many 
other complex, multilevel concepts in psychiatry the diagnostic cate- 
gories, because of unreliabiUty and subjectivity of conception, possess 
no standard criterion value. 

Five years of experience in applying the interpersonal system to 
clinical problems, plus the validating evidence from MMPI studies, 
have led to the conclusion that the relationships presented in Table 9 
provide a satisfactory functional definition of these six psychiatric 
categories. In a later section of this book eight chapters will be de- 
voted to the eight basic interpersonal diagnostic types. As we take up 
each adjustive-maladjustive mode, we shall review the Hterature per- 
taining to the psychiatric equivalents. At that time it will be suggested 
that the standard clinical definitions of these six Kraepelinian-type 
categories do involve interpersonal factors that tend to substantiate the 
relationships from our MMPI studies which are summarized in 
Table 9. 

The logic of operational definition, we recall, allows the scientist to 
define his concepts in terms of his measurements. Because of the use- 
fulness of psychiatric diagnostic terminology, and its implicit inter- 
personal connotations, it has seemed valuable to include it within the 
interpersonal diagnostic system. The relationships presented in Table 
9 therefore stand as operational definitions of the psychiatric terms 
concerned (at the designated level). In the subsequent chapters we 
shall be employing the terms "overconventional," "dependent," etc. 
When they are used it will be understood that they correspond to the 
respective psychiatric equivalents (at Level I-M or II-C) as indicated 
in Table 9. 


The Use of Standard Psychiatric Terms in Interpersonal Diagnosis 

The Kraepelinian-type categories are taken to be synonymous for 
the appropriate maladjustive types. According to this system hysteric 
is a synonym for over conventional personality; obsessive can be used 
interchangeably with masochistic personality . The Kraepelinian-type 
terms do not interchange with the adjustive types, but only with the 
maladjustive types. 

The new clinical terminology is employed in exactly the same man- 
ner as the interpersonal maladaptive categories. If a patient manifests 
extreme LM at Level I and extreme FG at Level II, he is designated in 
interpersonal terminology as an overconventional-distrustfjil personal- 
ity (numerical code = 74). For research purposes we employ the nu- 
merical code. If we wanted to communicate with a future therapist 
who is familiar with the interpersonal system we would probably use 
the interpersonal terms. If we wanted to communicate with a con- 
ventionally trained psychiatrist who is unfamiliar with the inter- 
personal language we would diagnose the patient hysterical-schizoid 
personality . 

Two additional considerations remain before concluding this chap- 
ter on interpersonal diagnosis. Table 9 presents the eight basic inter- 
personal types and lists many standard psychiatric categories which 
seem to relate to them. According to this system of diagnosis, phobics, 
neurasthenics, reactive depressives, and anxiety neurotics all can be ex- 
pected to manifest docile-dependent trends. When the diagnostic 
label is tied to the security operations displayed by the patient, some 
changes in meaning and accompanying paradoxes can be expected. It 
is possible for a patient to exhibit a phobic or anxiety symptom and 
not the dependent behavior that we expect to go along with it. Some- 
times a most aggressive or boastfully self-confident patient comes seek- 
ing psychiatric help for phobic complaints, i.e., irrational fears or anx- 
iety reactions. We diagnose, however, on the basis of character or 
symptomatic pressure, not on the basis of the symptom itself. The 
diagnosis summarizes the presenting operations. In cases of this sort we 
generally find that the symptoms are a result of some threat to the 
overt security operations. Sometimes the underlying operations reflect 
the interpersonal themes that go with the symptom. Thus the pa- 
tient who presents (at Levels I-M and II-C) as a counterphobic, self- 
satisfied person with a phobic symptom would be given the present- 
ing diagnosis of competitive or narcissistic personality {22 or 22). 
Examination of his "preconscious" behavior might reveal fearful 
and dependent themes (which we have seen to be related to phobic 


The symptom can be related to any level of personality, or it can 
simply reflect an environmental pressure which makes the overt se- 
curity operations inadequate or inappropriate. An example might be an 
acute anxiety attack suffered by a compulsive, managerial personality. 
If the multilevel profile involves "preconscious" passivity (e.g., 1 166), 
we might surmise that the symptom represents a leaking out of the 
underlying fear. If the multilevel pattern involves a solid four-layer 
edifice of strength and power (e.g., 1111), then the symptom would 
undoubtedly be a reaction to an environmental pressure for which his 
compulsive, managerial operations are inappropriate or inadequate. 
The pedantic, compulsive professor may be threatened by the loss of 
his job; or the bossy, self-made businessman may be threatened by the 
prospect of failure. Systematic diagnosis cannot, therefore, allow itself 
to be tied to description of symptom. The solution, we submit, is sys- 
tematically to describe and summarize behavior at the levels in which it 
is manifested. 

A final point merits comment. In looking over the list of standard 
psychiatric diagnostic terms to which we have given interpersonal re- 
definition, it will be noted that a few common categories are omitted. 
Among the terms which are left out of the list of maladjustments are 
such familiar labels as manic-depressive, paranoia, catatonia, etc. 

These have been excluded because there seems to be no typical 
interpersonal pattern associated with them. The essence of these dis- 
orders is an inconsistent behavior. The emphasis, on the contrary, 
seems to be on the changeability of behavior. This is obviously true of 
the hyphenated term manic-depressive — the variability is the essence 
of the personality. The term catatonic (which seems, by the way, 
vaguely defined and diminishing in popular usage) is generally 
described in variability terms. Paranoid, however, seems on the sur- 
face to have a most clear interpersonal meaning. It is used, in fact, 
synonymously with the distrustful, suspicious personality . While not 
everyone would agree, we have come to the conclusion that paranoid 
is one of the most loosely defined words in the psychiatric dictionary. 
A brief review of its connotations will reveal its protean and paradoxi- 
cal complexity. We have already mentioned its denotation of suspi- 
cion (FG). It also refers to delusions of grandeur (BC). Paranoids 
are generally associated with litigations, quarrelsomeness, or dangerous 
outbursts of aggression (DE). They often present themselves as 
pedantic and domineering (AP). They often claim to be hypernormal 
— denying pathology and weakness (NO). They are most frequently 
characterized by an obtuse, self-righteous, bland overconventionality, 
which Robert E. Harris has called "poignant naivete" (LM, JK). 

It is clear that there are a variety of specific interpersonal and 



symptomatic pressures that the so-called "paranoid" puts on the chni- 
cian. In fact, it seems that the essence of the term implies a complex, 
unapproachable, unstable personality. Many clinicians tend to sniff 
out the so-called "paranoid reaction" whenever they have the uneasy 
sense that multiplicity (and perhaps duplicity) of motive exists in the 
person they are dealing with. The essence of this diagnosis might be 
reduced to these factors: (1) complexity and variety at the level of 
presentation, (2) underlying hostility and distrust (which cannot be 
included in the diagnostic summaries of presenting operations), and 
(3) duplicity and/or self-deception (i.e., discrepancy between self- 
perception and view of self by others). These three criteria seem to 
suggest that the diagnosis of paranoia revolves around a certain pattern 
of variability (especially conflict between presenting operations versus 
underlying themes, and between the two types of presenting opera- 
tions [Level I-M versus Level II-C]). 

For these reasons the psychiatric terms paranoid, manic-depressive, 
and catatonic are not considered as denoting interpersonal patterns of 
presenting operations. They seem, instead, to p6int to certain phe- 
nomena of changeability, conflict, cyclical oscillation, and mispercep- 

The Diagnostic Continuum. In the preceding section we have 
cited evidence relating the interpersonal types to standard psychiatric 
diagnoses. The interpersonal types are on a continuum in such a way 
that neighboring behaviors are related, and behaviors opposite on the 
circle are considered to be negatively related. 

When we substitute psychiatric diagnostic terms for interpersonal 
categories we are suggesting that a diagnostic continuum exists. 

The advantages of such a continuum (if valid) are considerable. 
The process of diagnosis can be changed from a hit-or-miss pigeon- 
hole classification to a more systematic enterprise. The reliability and 
meaning of diagnosis can be increased. If one cHnician using the 
standard nosology calls a patient an hysteric, and a second labels him 
as phobic, a complete diagnostic "miss" must be registered. The use of 
the diagnostic continuum can clarify this situation. In this illustrative 
situation the two diagnosticians would be considered to be in fairly 
close agreement since they are just one unit off in their disagreement 
(since hysteric is one unit removed from phobic on the continuum). 

The value of this system depends, of course, on its validity. We 
have listed the diagnoses in a rough ordinal array. Does this make 
clinical, empirical sense? Are hysterics closer to phobics than they are 
to obsessives? 

The evidence from the Kaiser Foundation research seems to con- 
firm the hypothesis that they are. This research, however, is based on 


measurements which are shallow and obviously not as broad or deep 
as clinical impressions. 

The validity of this diagnostic continuum can be checked by the 
reader with clinical experience who can determine if the ordering of 
categories correlates with his diagnostic experience. 

The meaning of the diagnostic continuum can be broken down as 
follows: Psychopathic personalities are held to be closest to schizoids. 
They both share the alienation and isolation from conventional be- 
havior. The former are more active in their hostihty, the latter more 

Schizoid personalities are also close to obsessives. They both share 
a pessimistic, self-derogatory attitude. The former are more bitter and 
distrustful, the latter are more self-derogatory and worried. 

Obsessive personalities are also close to phobics. They both share a 
depressed, worried passivity. The former are more guilty, more aware 
of their emotions. The latter fail to recognize the emotional sources 
of their condition and are more concerned with symptoms external to 
their character structure. 

Phobic persoTialities are also close to hysterics. They both share a 
repressive, conventional facade. They both externahze and tend to be 
unaware of specific interpersonal problems. The former are more fear- 
ful and worried, the latter are more bland and unworried. 

The hysterical personality is also close to the psychosomatic adjust- 
ment. They both share conventional operations and claim to be un- 
worried and sound "psychologically." The former are more aware of 
some symptomatic "tension"; they present physical symptoms which 
are directly symbolic of underlying emotions. They are also relatively 
more passive and conciliatory. The psychosomatic personality em- 
phasizes more activity and hypernormal responsibility than the hys- 

The managerial and narcissistic personality types are not usually 
considered standard diagnostic categories and will therefore be 
omitted from these comparisons. 


1. AscH, p. The reliability of psychiatric diagnoses. /. abn. soc. Psychol., 1949, 44, 

2. DoERiNG, C. R. Reliability of observation of psychiatric and related characteristics. 
Amer. J. Orthophyschiat., 1934, 4, 249-57. 

3. Elfin, F. Specialists interpret the case of Harold Holzer. /. abn. soc. Psychol., 
1947,42, 99-111. 

4. Harris, R. E. Psychodiagnostic testing in psychiatry and psychosomatic medicine. 
In Recent advances in diagnostic psychological testing. Springfield, III.: Charles C. 
Thomas, 1950. 

5. Masserman, J. H., and H. T. Carmichael. Diagnosis and prognosis in psychiatry. 
/. Ment. Sci., 1938, S4, 893-946. 


The Variability Dimension of Personality: 
Theory and Variables 


The Kaiser Foundation research project works within the scope of 
two areas of personality — the interpersonal and variability dimensions. 
In Part II we presented the five levels at which we measure inter- 
personal behavior, and a multilevel system of interpersonal diagnosis 
was described. 

This section of the book presents an over-all view of the variability 
dimension and the Kaiser Foundation theory of variability. 

The variables by which we measure conflict and interlevel dis- 
crepancy are called variability indices. These are the variables of per- 
sonality organization — which relate behavior at different levels. Chap- 
ter 1 3 presents operational definitions of forty-eight indices of varia- 
bility. Some of these interlevel relationships are like classic psycho- 
analytic defense mechanisms. These indices are described and defined 
because we shall be employing them in the subsequent descriptions of 
clinical and diagnostic types. Their detailed description, validation, 
and clinical applicability will not be included in this book. 

Our present purpose is to outline a system of interpersonal diagnosis 
and the variables by which this is accomplished. Some of the research 
findings which involve variability indices (e.g., identification and mis- 
perception) are summarized in the diagnostic section (Chapters 14 
through 23). Other descriptions and validations of variability indices 
have been published in scientific journals (I, 2). 

The chapter to follow will, therefore, be restricted to a brief discus- 
sion of theory and a listing of the indices of variability. 


1. LaForge, R., T. Leary, H. Naboisek, H. Coffey, and M. Freedman. The inter- 
personal dimension of personality: II. An objective study of repression. /. Pers., 
1954, 23, 2, 129-53. 

2. Leary, T., and H. Coffey. The prediction of interpersonal behavior in group 
psychotherapy. Fsychodrama Group Psychother. Mojiogr., 1955, No. 28. 



The Indices of Variability 

The preceding pages of this book — Chapters 1 through 12 — have con- 
sidered the interpersonal dimension of personality. We have seen that 
security operations can be classified in terms of sixteen variables. We 
have also considered the fact that interpersonal behavior exists at dif- 
ferent levels and that these levels may be defined in terms of the source 
and nature of their expression. 

The interpersonal dimension has five levels. A circular continuum 
of variables is employed for all measures of emotional behavior. We 
have dealt in some detail with the measurements, meaning, and func- 
tion of these five levels. In presenting this material, the levels were 
considered separately. The point was made repeatedly that the data 
from any level are most useful in relation to all the other levels. But so 
far we have concentrated on statements about each level in isolation, 
because we have had no systematic way of dealing with the dynamic 
interplay among levels. The time has come to discuss these relation- 
ships. We are going to fit together four^ discrete parts of personality 
structure in order to build a systematic theory of personality organi- 
zation. We are going to study the integration of the over-all personal- 
ity. The shift in reference we make here is important to note. Up 
until now we have dealt with unilevel data. Now we are going to 
study multilevel phenomena — the dynamics of organization. It is 
necessary, at this point, to distinguish between statements that refer to 
security operations and to the arrangement of interpersonal variables 
at a single level or area (i.e., the circle) and those which refer to the 
differences among the circles. The former can be called class (i.e., 
unilevel) statements and the latter relationship (i.e., multilevel) state- 

When the interpersonal behavior of an individual at one level of 
personality is classified, the resulting data might take the form of the 
following: "The patient complains to the therapist." "He attacks the 

^ Level IV not included. 



other group members." These are class statements. They refer to one 
area of security operations — Level I Communications. A patient says, 
"I like my therapist and the other group patients." These are also class 
statements. They refer to another area of behavior — Level II, Con- 
scious Descriptions. 

When behavior at one level is compared to behavior at another 
level, relationship statements are being made. The discrepancies be- 
tween levels become the focus. We note, for example, that a patient 
consciously reports himself as friendly, although his behavior as rated 
by others is hostile. We might say that this patient misperceives his 
hostile behavior. The word Tmsperceives is a relationship term since it 
compares two levels of observation — the subjective and objective 
view of the patient's behavior. 

When we shift from sentences about a single level to sentences 
about the differences between levels, a new set of concepts is involved. 
The term misperceives is an example of such a concept, A new vocab- 
ulary and syntax come into play. This comes under the heading of 
the "Logic of Levels." The distinction between class and relationship 
statements must be kept clear or faulty conceptualization will result. 

What Is the Variability Dimension? 

We are dealing here with an entirely new and different type of 
data — the relationships among the levels of personality. We designate 
this as the variability dimension of personality. This is a most impor- 
tant aspect of behavior. Variability has classically been the stumbling 
block in the development of personality theory. Every systematic 
treatment of human nature has had to labor with the perverse incon- 
sistency of behavior. It has never been difficult for theorists to invent 
typologies and variables of emotions. The trouble has always come 
when the elusive human subject begins to demonstrate his protean 

The first theoretical lever which succeeded in moving this obstacle 
was provided by the theory of unconscious motivation. When Sig- 
mund Freud defined the multilevel nature of personality, he offered the 
first systematic explanation of conflicts, ambivalence, and incon- 
sistency. The essence of a dynamic psychology is variability. The 
great advantage of a depth theory is that it explains the puzzling com- 
plexity and contradiction inherent in human behavior. The psycho- 
analytic theory of personality and of neurosis is defined in terms of 
interlevel conflict, i.e., variability. There are certain motives at one 
level and certain other motives at another. Their interplay spells out 
the organization of personality. When Freud presents his great tri- 


partite division of character into ego, superego, and id, he is mainly 
concerned with conceptualizing the multilevel variability of behavior. 
VariabiUty is involved in almost every aspect of personality that we 
study. Conflict, ambivalence, defense mechanisms, growth, regression, 
change, improvement-in-therapy — all these phenomena have the basic 
factor in common — one unit of measure varies in relation to another. 
It is a major thesis of this book that all change phenomena are, to a 
certain extent, functions of a general rigidity-oscillation factor. This 
factor is measurable and predictable in terms of the time and the 
amount of variability. Some human beings are more variable, some are 
less. Some express variation between certain areas of their personality; 
others express it in different areas. The amount and kind of variability 
is a most significant variable of human behavior. It defines the type 
and intensity of conflict. It determines the tendency to change or to 
maintain a rigid adjustment. It becomes a key variable in the diag- 
nostic and prognostic formulas through which we conceptualize hu- 
man personality. 

Structural, Temporal, and Situational Variability 

In considering the variability factor it is useful to make the follow- 
ing distinctions between structural, temporal, and situational variabil- 

Structural variability refers to differences among the levels of per- 
sonality. It is well known that drastic discrepancies and inconsistencies 
develop when we compare the conscious self-description with be- 
havioral or symbolic expressions. The subject who presents himself as 
a warm-hearted, tender soul may produce dreams or fantasies which 
are bitterly murderous. Social interactions, as observed by others, may 
be quite different from the subject's own view of them. 

Temporal variability refers to inconsistencies in the same level of 
behavior over a time span. Time inevitably brings changes, great or 
small. Many subjects show marked cyclical swings of mood or ac- 
tion. The interpersonal behavior of an individual generally mutates 
as he moves from age 13 to 31. The temporal changes we study in 
psychiatric patients are called spontaneous remissions, therapeutic re- 
coveries, psychotic episodes, and the like. 

Situational variability refers to differences in cultural and environ- 
mental factors. The man who is a lion at home may be a lamb in the 
office. Reactions often vary according to the sex, age, and cultural 
status of the "other one" with whom the subject is dealing. 

The Kaiser Foundation research project is studying the hypothesis 
that all of these types of changes are related to the same variability 


factor. To distinguish between them may be an artifact, operationally- 
useful in the light of the scientific manageability of change phenomena. 
It is difficult enough to measure interpersonal behavior at one time 
and at one level. To study variation it is necessary to have two sets of 
data on the same subject which we compare. Three distinct operations 
are required. We must measure one set of behaviors, then the other, 
and finally, the discrepancy or change index. We can isolate tem- 
poral variation by holding constant the level and the situation from 
which the data come, and measuring the change over time. Repeating 
the same personality test on a control patient might be an example of 
this type of variation. Observing in a group therapy session a patient's 
successive reactions to a nurturant therapist and a competitive fellow 
patient would illustrate a change in the interpersonal situation often 
leading to variance in the subject's responses. If we establish the kind 
of variability in one of these classes (e.g., in personality structure) we 
can make probability predictions as to the kind of variation to be ex- 
pected in another class. 

Comparison Between the Interpersonal 
and the Variability Dimension 

The network of relationships of the various parts of the system — 
and that is what we deal with here — obviously tells us something about 
the organization of that system. Interlevel discrepancies are therefore 
indices of organization. They tell us about the agreement or conflict 
between the various levels of behavior. They tell us not how the sub- 
ject relates to his environment, but rather how the different areas of 
his personality relate to each other. In the interpersonal dimension 
we study different phenomena and employ different variables. For ex- 
ample, the subject's Level I profile is measured in terms of the sixteen 
interpersonal variables, summarizing his actual relationships with other 
people. The Level II profile employs the same sixteen variables to 
summarize his consciously described relationships with other people. 
When we compare the Level II and III profiles we get a discrepancy 
score. We move into a new dimension and must employ a new set of 

These variables are not interpersonal, but intrapersonal. Projec- 
tion and suppression are terms used to describe certain kinds of rela- 
tionship between levels of personality data. Notice that they are not 
directly interpersonal; one does not project or suppress another per- 
son. He projects or suppresses his own private motivation. These re- 
lationship variables refer not to his social relationships but to the rela- 
tionships which hold between the areas of his own behavior. Similarly, 
the rating of amount of interlevel discrepancy must be distinguished 


from an interpersonal rating. We do not use the terms rigidity and 
conflict to describe what one person does to another — we employ them 
to describe the tightness, looseness, consistency, or ambivalence among 
the levels of personality. 

Thus, we have introduced into the interpersonal system not just a 
new variable, but a new category of classification. This is called the 
variability dimension of personality. A new order of measurement is 
involved which taps all of the variability phenomena of human be- 
havior: similarity-difference, change, discrepancy, conflict. There are, 
as we have seen, many types of changes that show up in human be- 
havior — those due to time, situation, and chance, as well as those due 
to lawful inconsistencies among the levels of personality. The vari- 
ability dimension is a formal aspect of personality as opposed to the 
more empirical procedures by which we measure interpersonal be- 
havior. Its variables are determined not from empirical observation, 
but from logical procedures. 

In developing the interpersonal variables, we began with actual 
interpersonal behavior. We collected emotional data of all kinds and 
then developed a system which best reflected the varieties of inter- 
personal purpose. In developing the levels of personality the same em- 
pirical technique was followed. The diff^erent sources of data were 
examined and combined into the four levels. But in determining the 
measurement and conceptual units for the variability dimension, a dif- 
ferent solution is involved. We are not dealing with actual human be- 
havior, but with indices of change — changes in the scores from the 
interpersonal dimension. The variables of the variability dimension are 
not units of the subject's behavior, but of the scientists' behavior, for 
it is the scientist who performs the operations.^ 

This is a point worth stressing. The number and kind of inter- 
personal variables were limited and determined only by our observa- 
tions of what individuals do to each other in their social interaction. 
The number of levels was similarly determined by empirical evidence. 
A certain flexibility in the selection of variables and levels does exist 
because any scientist has the right to increase or diminish the number 
of categories by which he classifies behavior. In setting up empirical 
categories, some room for interpretive judgment is allowed — the em- 
pirical data guides, but does not dictate. Once the number of levels is 
determined, however, the system becomes "set." When the scientist 

^ To be more precise we should say that the variability dimension is twice removed 
from behavior, and the interpersonal dimension is once removed. The patient does 
something; then the scientist categorizes or measures it. The interpersonal data ob- 
tained in this way are once removed from the subject's behavior. Then the scientist 
goes on to compare the different levels of areas of interpersonal data. These formal, 
analytic operations are thus twice removed from raw behavior. 



goes on to compare the differences between the levels he has selected, 
there is no longer any freedom for interpretation or creative choice. 
The logic of levels takes over and dictates the range and nature of the 
interlevel discrepancies. 

An empirical system for measuring multilevel behavior leads, then, 
to a somewhat new theory of personality organization. A different 
conception of conflict is defined, not in terms of the interplay between 
postulated forces, but in terms of the discrepancies between measures. 

The Indices of Variability 

The relationships between areas of personality are called variability 
indices. We have already discussed their general similarity to the con- 
ception of (but not the clinical use of) Freudian defense mechanisms. 
These indices comprise one aspect of the variability dimension of per- 
sonality. They reflect the stabihty or variation existing among the 
levels of personality at one point in time. They are to be distinguished 
from other kinds of change phenomena included in the variability di- 
mension, such as modulations over time (which includes change in 
psychotherapy) or variability due to differences in the cultural situa- 
tion. We deal here with indices of interlevel conflicts and con- 

The term variability index has been assigned to this kind of varia- 
tion for the following reason. This is a rather neutral term. It is in- 
tended to point out that we are not dealing with mechanisms or even 
with behaviors, but rather with formal, comparative operations. These 
discrepancy relationships do not "do anything." It is risky to mechan- 
ize or humanize them. There are no body organs or neural centers for 
repression or suppression. 

From the empirical point of view, there is simply behavior at dif- 
ferent, discriminable levels of expression. There are measured rela- 
tionships between these levels. In the present insecure state of our 
knowledge it seems safest to call them indices. But indices of what? 
To answer this question is to produce a theory of personality organiza- 

Variability index is, we expect, a temporary holding term which 
can be replaced by a more dynamic term whenever the nature of the 
dynamic principle is determined. In the meantime, it seems to express 
exactly what we know to be true about the interlevel relationships. It 
tells us how stable or variant these relationships are. If the themes of 
Level II parallel those of Level III, then the variability index is low. 
The aggression, let us say, of one level is repeated at the other. If the 
two levels are discrepant — if, for example, the aggression at Level II 
changes to docile cooperativeness at Level III — a high variability index 


is obtained. Variation between the levels is present. There is another 
and more important aspect of variability. There is evidence (see Ap- 
pendix 3 ) that the more stable the organization of personality — that is, 
the more the data from Levels II, III, and I tend to repeat the same 
themes — the less variation we can expect in the personality organiza- 
tion over time. Conversely, the more conflict or oscillation among the 
levels of personality, the more change we can predict will take place 
in the future; and this includes change in therapy. These findings 
make the term variability index doubly appropriate. While we can- 
not, at this point, say that interlevel discrepancies possess the dynamic 
qualities of pushing toward equilibrium, we can say that they refer to 
structural stability of personality (this by definition), and they predict 
the degree of stability of personality organization to be expected in 
the future. They give us, first of all, an index of systematic variation 
in the personality structure at the time of evaluation, and they point 
out the direction and amount of change to be expected over future 

The Function of Variability Indices 

The interpersonal system does not assign a function to these inter- 
level discrepancies. Behavior at all levels is seen as having one basic 
function to ward off survival anxiety. The discrepancies or conflicts 
between levels are seen as another dimension of conception which con- 
cerns the psychologist's behavior. It is the psychologist who measures 
the discrepancy or conflict between the two levels of the patient's be- 
havior. We assign, for semantic convenience and heuristic necessity, 
conceptual titles to the important discrepancies between the levels of 
the subject's behavior. We do not, however, assign functions to them. 
The only assumption upon which an empirical theory of personaUty 
need be based is the premise of survival anxiety. The only function 
we assign to behavior is the maintenance of security and the diminish- 
ing of anxiety. The indices of diff^erence, ambivalence, or conflict 
among the varieties of behavior do not seem to require the postulation 
of additional functions. 

Closely connected to this question of the function of defense mech- 
anisms (or variability indices) is another issue which has received con- 
siderable attention in the recent literature. This involves the differen- 
tiation between adaptive and defensive functions. Are defense mech- 
anisms pathological and neurotic, or can they sometimes be construc- 
tive? Fenichel, for example, places all "successful defenses" under the 
heading of sublimation, and describes "unsuccessful defenses which 
necessitate a repetition or perpetuation of the warding-off process to 
prevent the eruption of the warded-off impulses." Mowrer has dis- 


tinguished between the mechanisms used in development and those 
used in defense. This issue of the adaptability or pathology attached 
to the discrepancies of concordance and conflict in personality is in- 
variably complicated by value judgments (e.g., what is adaptive?) and 
tlieoretical assumptions about the function of defense mechanisms. 
These are legitimate questions from the standpoint of the psycho- 
analytic approach and deserve the attention they have received. 

From the position of the interpersonal system, this issue could be in- 
terpreted as follows: (1) the variabihty indices have no function; (2) 
discrepancy or conflict between levels cannot be assigned an adaptive 
or maladjustive value by definition but must be interpreted as part of 
the total personality picture. The level and amount of the conflict and 
its relationship to the over-all character structure determine the posi- 
tive or negative interpretation. 

A not infrequent clinical misinterpretation of psychoanalytic 
theory implies that defense mechanisms are negative or neurotic proc- 
esses. This is, indeed, one reason which supports the use of the more 
neutral term variability index for the interlevel conflicts. A discrep- 
ancy between conscious self-description and "preconscious" fantasy 
(which we shall designate repression) should not necessarily be con- 
sidered unhealthy. If the Level II self-image is one-sided and the "pre- 
conscious" fantasy a moderate balance in the opposite direction, the 
conflict might well designate an adaptive equilibrium. If the patient is 
markedly disidentified with his father, the adjustive aspect of this dis- 
crepancy would certainly depend somewhat on the kind of motives 
attributed to self, to father, and to others. 

In the subsequent pages we shall be considering several variability 
indices which have been given names of psychoanalytic defense mech- 
anisms where these seemed to fit the nature of the conflict. In order 
to understand the meaning and use of these indices in the interpersonal 
system, it is essential that two points be kept in mind: (1) These con- 
cepts are not mechanisms or dynamisms, but rather numerical indices 
of interlevel variation; as such they have no function. (2) They have 
no a priori value-loading as far as adjustment and maladjustment are 
concerned; they can describe flexibility and healthy ambivalence, or 
they can indicate pathological rigidity or maladaptive oscillation. 

Two Interpretations of Variability 

When we obtain the variability indices among the levels of per- 
sonality, two interpretations of the resulting variation can be made — 
both of theoretical and practical interest. We can concentrate on 
what the variation is, or we can focus on how Tmich. The first tells us 
that the individual represses so much hostility or misperceives this 


much passivity. The second way of handhng variability indices is to 
disregard the content of the interpersonal themes, study the pattern of 
variability for all the discrepancy relationships, and simply determine 
how variable this person is in over-all terms. This focuses on the 
amount of variability. We can then make such statements as, "This 
patient is extremely conflicted and variable, being two sigmas above 
the mean." 

We have seen that the logic of levels determines the kind of rela- 
tionships among levels. We have developed a system in which there 
are eight general levels and areas of personality. Therefore, when 
we ask the question, "What are the relationships among the areas of 
personality?" the answer is already settled for us. They are the rela- 
tionships among these eight areas — the discrepancies which occur 
when we compare each level or sublevel with every other level. 
Formally, then, there are as many relationships or variability indices 
as there are permutations among the areas. 

Figure 31 presents these eight areas. Each circle represents a dis- 
crete area of personality data. The lines joining the circles represent 
the interlevel or interarea discrepancy indices determined by the logic 
of levels. They comprise the network of variability indices which 
link the parts of personality structure into an organized totality. 
Validation of these indices is beyond the scope of this book. Thirteen 
of these relationships — those most relevant to current theory and 
clinical practice — are defined in this chapter. 

The Record Booklet for Interpersonal Diagnosis of Personality 
(Appendix 4, Figure 61) provides a simplified method for measuring 
discrepancy indices and for plotting them in diagrammatic summary 

The next task is to determine the meaning of these interlevel rela- 
tionships. The subtractive procedures, it will be recalled, indicate 
the kind and amount of interpersonal behavior in one area that is pres- 
ent in another area. Giving names to these relationships is, in one way, 
the simplest problem of all. The term which best mirrors the rela- 
tionship is selected and operationally defined in terms of the cross-level 
subtraction. A procedure of this sort satisfies all the logical require- 
ments, but the reader is likely to remain unsatisfied and to ask the 
further questions, "This is all very well, but what do they mean? 
What is their functional value? What do they predict?" 

A scientific system can be objective and logically virtuous and still 
have no function except perhaps to entertain the originator. It would 
be possible to assign very impressive terms to the interlevel relation- 
ships, calling this one "repression" and that one "displacement" and a 
third "introjection," etc., until the long list of relationships (or the 





imagination) is exhausted. The resulting nomenclature would be 
logically consistent and objective (since all the terms would be opera- 
tionally defined), but, what is rather unfortunate, it would be quite 
irrelevant. In developing a system of personality, the first problem 
is the selection of the categories. The next is to validate them, that is, 
to relate them to other independent and relevant variables, to harness 
them to functionally useful predictions. The three criteria for effec- 
tive research, we recall, are objective measurement, logical analysis, 
and, far from the least important, functional relevance. 

If it were feasible to list and label all the possible interlevel relation- 
ships, the next task would be to validate them against functional 
criteria. Such labeling and validating would place an enormous drain 
on inventive imagination, research resources, and reader endurance 
alike. At this point, we shall attempt to define twelve generic varia- 
bility indices, and then list forty-eight specific indices which fall into 
the twelve broader categories. 

The twelve variability indices about to be defined have been chosen 
because they appear to possess the most clinical meaning, functional 
value, and theoretical implication. As we begin this exercise in the 
mathematics of personality, it is well to keep in mind the formal or 
logical aspect of the task. In one sense, it is not absolutely necessary 
to develop a notational system for linking up the levels of personality. 
In our diagnostic procedures we could conceivably just present the 
interpersonal behavior at all levels. We would indicate that the sub- 
ject is hostile at Level I, claims to be docile at Level II, describes his 
father as autocratic at Level II Other, etc. The language of variability 
allows us to relate these areas or levels of personality. It allows us to 
define systematically the dynamic network which links up the de- 
scribed areas. This is a great convenience. Like any formal, notational 
device, the language of variability makes possible concise, precise sum- 
maries of conflict, concordance, discrepancy, etc. 

Operatioiial Definition of the Variability Indices 

There are twelve generic variability indices to be defined and 
validated in this chapter. Most of these generic discrepancies have 
several subdivisions which are specific indices referring to the impor- 
tant familial figures to which the subject is related. Thus, under the 
generic index conscious identification there are four specific indices 
referring to identifications with father, mother, spouse, and therapist. 
Table 10 presents the twelve generic variability indices and indicates 
the subvarieties which are subsumed under this general title. 

In the left column of Table 10 are listed the most familiar titles of 
the twelve generic variability indices. In the right-hand column are 



Informal Listing of the T'welve Generic Variability Indices 

Code Number of the Specific Variability 
Title of Variability Indices Indices Subsumed Under this General Title 

Role Coincidence 11 SO 

Interpersonal Perception 12 SS, 12 OO 

Conscious Identification 22 SM, 22 SF, 22 SSp, 22 ST 

Equation 22 MF, 22 MSp, 22 FSp, 22 MT, 22 FT, 22 SpT 

Repression 23 SH 

Cross-level Identification 23 MH, 23 FH, 23 SpH, 23 TH, 23 SO, 23 SM, 

23 SF, 23 SSp 

Conscious-"Preconscious" Fusion 23 MM, 23 FF, 23 SpSp 

Displacement 23 A'lF, 23 FM, Z3SpM, 23 SpF 

23 M Sp, 23 F Sp, 23 TM, 23 TF, 23 TSp 

"Preconscious" Identification 33 HM, 33 HF, 33 HSp, 33 HO 

Self-Acceptance 25 IS 

Conscious Idealization 25 IM, 25 IF, 25 ISp, 25 IT 

"Preconscious" Idealization 35 IH, 35 IM, 35 IF, 35lSp, 35 lO 

noted the code designations of the specific variability indices which 
are the subvarieties of the generic indices. It will be observed that 
there are forty-eight of these specific indices. All of these will be 
operationally defined in the subsequent pages. 

It will be noted that many of these generic variability indices have 
been given the names of classical psychoanalytic defense mechanisms. 
Although borrowing terminology from another theory has its risks, 
we have ventured to employ the familiar terms wherever they seem to 
fit the general nature of the discrepancy concerned. In this manner 
we have sought to avoid the proliferation of novel terms and the 
idiosyncratic "timid neologisms" which Egon Brunswick has deplored. 
Several of the discrepancy indices, however, involve reflex inter- 
personal communications — a level of personality which has not been 
isolated by the psychoanalytic theory. This has necessitated the intro- 
duction of some new terminology — or in Freudian language, some 
new "defense mechanisms." 

Confusion between these variability indices and psychoanalytic 
mechanisms of the same name may be avoided if the reader keeps in 
mind the operational definition of each index. It may be helpful to 
present a diagrammatic operational definition of these twelve vari- 
ability indices. They are defined by the amount of discrepancy be- 
tween levels or areas of personality as illustrated in Figure 31. 

Several points require comment. First, it will be noted that the re- 
lationships of Level IV (the level of the unexpressed unconscious) to 
the other levels are not included. No data are available for this area of 
personality. It must also be noted that only one circle is presented for 


each area of "other" behavior.^ In practice, there are several "others" 
who are always included in the personality diagram. As we recall 
from the chapter on Level II (Chapter 8), the conscious view of 
mother, father, and spouse (and, where possible, the therapist) is 
routinely included in the personality pattern. Similarly, in scoring 
fantasy material, we separate the themes attributed to father, mother, 
and cross-sex figures. These specialized circles are not all included 
in Figure 31. We have included an extra Level II Other circle, labeled 
"father," to illustrate the variability indices of Familial equation, de- 
fined as the process of consciously ascribing similarities or differences 
to various family members or describing nonmembers (such as the 
therapist) as being like or unlike family members. If a patient de- 
scribes hij therapist in the same way that he describes his father, the 
two indices will show little or no discrepancy. We would be able to 
say, "The patient consciously equates his therapist with his father." 
We have also included an extra Level III Other circle to illustrate 
the variability indices of displace?nent, which is defined as the process 
of consciously ascribing to one "other" (e.g., father) the interpersonal 
traits which are preconsciously assigned to another "other" (e.g., 
mother) . 

Operational Definition of Forty-eight Specific Variability hidices 

It was mentioned above that the twelve generic variability indices 
subdivide into forty-eight specific variability indices. If we consider 
all the permutations and combinations of interrelatedness among the 
levels and the personages at each level, a list of variability indices sev- 
eral times forty-eight would be obtained. The forty-eight indices now 
to be defined were selected on the basis of the theoretical and clinical 
meaningfulness. The plan of exposition is as follows: We shall first 
present an operational definition for each of the forty-eight variability 
indices and a formal title for the high and a low discrepancy for each. 

The listing, coding, formal designation, and operational definition of 
each variability index is contained in Table 1 1 , The key to the num- 
bers and letters employed in coding the variabihty indices is presented 
in Table 12. 

The Coding of the Variability Indices 

The first column in Table 1 1 gives the code number of the variabil- 
ity index. The code number is a simple, straightforward notational 
device which summarizes exactly what discrepancy is involved in this 
index. Every code number for a variability index comprises four 

^ An exception to this statement— two circles are included in Figure 31 for Level II 
Other and Level III Other to illustrate the indices of equation and displacement. 



Operational Definition of Forty-eight Indices of Variation 

Low Discrepancy 

High Discrepancy 

Between the Two 

Between the Two 


Measures Is Called: 

Measures Is Called: 

11 SO 

Role coincidence 

Role reciprocity 

12 SS 



12 GO 


Other misperception 

22 SM 

Conscious identification 

Conscious disiden- 


tification (maternal) 

22 SF 

Conscious identification 

Conscious disiden- 


tification (paternal) 


Conscious identification 

Conscious disiden- 


tification (spouse) 

22 ST 

Conscious identification 

Conscious disidentifi- 

with therapist 

cation with therapist 

22 MT 





22 FT 










22 MF 















23 SH 





23 MH 

Cross-level identifica- 

Cross-level disidentifi- 

tion (maternal) 

cation (maternal) 

23 FH 

Cross-level identifica- 

Cross-level disidentifi- 

tion (paternal) 

cation (paternal) 


Cross-level identifica- 

Cross-level disidentifi- 

tion (spouse) 

cation (spouse) 

23 TH 

Cross-level identifica- 

Cross-level disidentifi- 

tion (therapist) 

cation (therapist) 

23 SO 

Cross-level identifica- 

Cross-level disidentifi- 

tion (other) 

cation (other) 

23 SM 

Cross-level identifica- 

Cross-level disidentifi- 

tion (maternal) 

cation (maternal) 

23 SF 

Cross-level identifica- 

Cross-level disidentifi- 

tion (paternal) 

cation (paternal) 


Cross-level identifica- 

Cross-level disidentifi- 

tion (cross-sex) 

cation (cross-sex) 

23 MM 

Fusion (maternal) 

Diffusion (maternal) 

This Variability Index 
Is Operationally De- 
fined by the Discrep- 
ancy Between: 

Level I self vs. Level I 

other (specialized or 

Level I self vs. 

Level II self 
Level I other vs. 

Level II other 
Level II self vs. 

Level II mother 
Level II self vs. 

Level II father 
Level II self vs. 

Level II spouse 
Level II self vs. 

Level II therapist 
Level II mother vs. 

Level II therapist 
Level II father vs. 

Level II therapist 
Level II spouse vs. 

Level II therapist 
Level II mother vs. 

Level II father 
Level II mother vs. 

Level II spouse 
Level II father vs. 

Level II spouse 
Level II self vs. 

Level III hero 
Level II mother vs. 

Level III hero 
Level II father vs. 

Level III hero 
Level II spouse vs. 

Level III hero 
Level II therapist vs. 

Level III hero 
Level II self vs. 

Level III other 
Level II self vs. Level 

III maternal images 
Level II self vs. Level 

III paternal image 
Level II self vs. Level 

III cross-sex images 
Level II mother vs. 

Level III maternal 




TABLE 11— Continued 
Operational Definition of Forty -eight Indices of Variation 


Low Discrepancy 
Between the Two 
Measures Is Called: 

23 FF 

Fusion (paternal) 

23 SpSp 

Fusion (cross-sex) 

23 MF 


23 FM 












23 TF 


23TSp Displacement 

33 HM 

33 HF 
33 HO 
25 IS 
25 IM 
25 IF 
25 IT 

"Preconscious" identi- 
fication (maternal) 

"Preconscious" identi- 
fication (paternal) 

"Preconscious" identi- 
fication (cross-sex) 

"Preconscious" identi- 
fication (total) 


Maternal idealization 
Paternal idealization 
Spouse idealization 
Therapist idealization 

High Discrepancy 
Between the Two 
Measures Is Called: 

Diffusion (paternal) 
Diffusion (cross-sex) 
Cross-level diffusion 

Cross-level diffusion 
Cross-level diffusion 

Cross-level diffusion 

Cross-level diffusion 

Cross-level diffusion 

Cross-level diffusion 

Cross-level diffusion 

Cross-level diffusion 

"Preconscious" disiden- 
tification (maternal) 

"Preconscious" disiden- 
tification (paternal) 

"Preconscious" disiden- 
tification (cross-sex) 

"Preconscious" disiden- 
tification (total) 

Self -rejection 

Maternal devaluation 
Paternal devaluation 
Spouse devaluation 
Therapist devaluation 

This Variability Index 
Is Operationally De- 
fined by the Discrep- 
ancy Between: 

Level II father vs. Level 

III paternal image 
Level II spouse vs. Level 

III cross-sex image 
Level II mother vs. 

Level III paternal 

Level II father vs. Level 

III maternal image 
Level II spouse vs. 

Level III maternal 

Level II spouse vs. 

Level III paternal 

Level II mother vs. 

Level III cross-sex 

Level II father vs. 

Level III cross-sex 

Level II therapist vs. 

Level III maternal 

Level II therapist vs. 

Level III paternal 

Level II therapist vs. 

Level III cross-sex 

Level III hero vs. 

Level III maternal 

Level III hero vs. Level 

III paternal image 
Level III hero vs. Level 

III cross-sex image 
Level III hero vs. Level 

III total other 
Level V ideal vs. 

Level II self 
Level V ideal vs. 

Level II mother 
Level V ideal vs. 

Level II father 
Level V ideal vs. 

Level II spouse 
Level V ideal vs. 

Level II therapist 



TABLE 11-Continued 
Operational Definition of Forty-eight Indices of Variation 


35 IH 

35 IM 
35 IF 
35 lO 

Low Discrepancy 
Between the Two 
Measures Is Called: 

"Preconscious" hero 

"Preconscious" mater- 
nal idealization 

"Preconscious" paternal 

"Preconscious" cross- 
sex idealization 

"Preconscious" other 

High Discrepancy 
Between the Two 
Measures Is Called: 

"Preconscious" hero 

"Preconscious" mater- 
nal devaluation 

"Preconscious" paternal 

"Preconscious" cross- 
sex devaluation 

"Preconscious" other 

This Variability Index 
Is Operationally De- 
fined by the Discrep- 
ancy Between: 

Level V ideal vs. 

Ill hero 
Level V ideal vs. Level 

III maternal images 
Level V ideal vs. Level 

III paternal images 
Level V ideal vs. Level 

III cross-sex images 
Level V ideal vs. Level 

III other 


Key to Numbers and Letters Employed in Coding Variability Indices 

Number Codes Letter Codes 

1 = Level I S = Self (i.e. the subject) 

2 = Level II M = Mother (the subject's own mother 

3 = Level III is at Level II and the maternal im- 
5 = Level V age at Level III) 

F = Father or paternal image 

Sp = Spouse (if at Level 'II) or cross-sex 

figures (in Level III fantasies) 
T = Therapist 
I = Ego ideal from Level V 

elements: two arable numbers and two letters. The numbers indicate 
which levels are being compared. Thus, "12" means that behavior at 
Level I is being compared with behavior at Level II; "22" indicates 
that the comparison is between two different scores at Level II. The 
two letters refer to the respective personages at each level that are 
being compared. Thus, "12 SS" indicates that the subject's own be- 
havior seen by others (Level I) and that seen by self (Level II) are 
being compared. The coding "23 FM" indicates that the subject's 
conscious description (Level II) of his father is being compared with 
the summed maternal images for Level III. 

It will be noted that any code number is actually a formula sum- 
mary of the processes involved in obtaining the index and is thus an 
abbreviated operational definition of the index. Since dozens of varia- 
bility indices are obtained for each patient studied in the Kaiser Foun- 
dation project, the codings allow a numerical filing system for variabil- 
ity data. 



The Titles of Variability Indices 

For each variability index, there are two titles which refer to high 
or low discrepancies between the two levels or personages involved. 
If a patient's self-description is close to his observed behavior (i.e., a 
low discrepancy) the first, or positive, designation "self-perception" is 
employed. If the two levels are far apart (i.e., a high discrepancy) 
then the second, or negative, designation "self-deception" is used. 

The specific procedure for measuring the variability indices will be 
presented below. It will suffice here to say that if a discrepancy be- 
tween two levels is below the mean of the normative sample, the 
positive term (second column in Table 11) is assigned. If the dis- 
crepancy is above the mean, the negative term (third column) is em- 

The Operational Definitions of Variability Indices 

The fourth column in Table 1 1 contains the operational definition 
of each of the forty-eight variability indices. This involves simply the 
specific designation of the levels and personages being compared. 

Methodology for Measuring Variability Indices 

The Kaiser Foundation psychology research project has devoted 
several years to the development of methods for measuring the kind 
and amount of variability between levels of personality and the kind 
and amount of variability between two tests of the same level ad- 
ministered at different times. The former are structural variability 
indices; the latter are called ternporal variability indices. 

A main criterion for an effective discrepancy measure is the re- 
flection of changes in line with the general meaning of the inter- 
personal circle. Thus a large numerical discrepancy between two 
levels or between the same level tapped in pretherapy and posttherapy 
tests should designate an extreme change in interpersonal behavior, 
e.g., from submission to dominance. 

One method of assessing discrepancy involved measuring the linear 
distance between the two scores in centimeters. This has the ad- 
vantage of directness and simplicity. It had the overweighing disad- 
vantage of doing violence to the concept of the circle. A large centi- 
meter difference was deceptive where the pre- and posttherapy scores 
were far from the center of the circle. Both scores could be in the 
same octant and involve a similar extreme interpersonal behavior (e.g., 
sadism) for which the centimeter distance index would be very large. 

The development of the numerical diagnostic codes (see Chapter 
12) made possible improved methods of measuring change.^ In the nu- 

* The remainder of this chapter was written by Joan S. LaForge. 


merical code system every score falling in the same octant is assigned 
the same code category. Thus a simple subtraction process yields a 
discrepancy estimate. If the patient is a 5 before therapy and a 2 after 
therapy, he has changed three units (i.e., 5 — 2 = 3). At first a crude, 
intuitive discrepancy system was established. Arbitrary values were 
assigned to the various differences. The comparison between the ex- 
treme and the moderate scores (e.g., 74) became an insoluble problem 
because there was no provision for assigning discriminatory weights. 



8 7 RED 

Figure 32. Model Employed To Determine Summary Points on the Diagnostic 
Grid and To Calculate Horizontal and Vertical Discrepancy Values. 



The attempt was then made to estabHsh a set of points on the circle 
to represent the eight octants at two intensity levels, a total of sixteen 
points. Any point falling in an octant is then considered to fall always 
at one point. A model establishing these points was derived in the fol- 
lowing way. The plane was divided into two areas (inside and outside 
areas), one standard deviation from the center point being the di- 
viding line. Those points in the inside area were considered to be of 
moderate intensity. The center of mass of each pie-shaped area was 
taken as the location of the representative point and derived from the 

where x and y are points along the abscissa and ordinate, and s is the 
area, integrated over the region R. Once this collection of eight points 
was established, the problem of finding a point-representation for the 
extreme intensity of each octant was a little more difficult. There can 
be no center of mass because each outer area is infinite. 

At this juncture, consideration of the meaning of relative discrep- 
ancies was taken into account. Clearly a change in the same octant 
should be less than the change, even at the least intensity, between two 
octants. With this principle in mind, a set of points was arbitrarily 
selected to represent the extreme intensities for the eight octants, main- 
taining, for example, that a red 1 -black 1 discrepancy be a little less 
than a black i-black 2 discrepancy. 

The intersection of each point with the x (hostility-affiliation) and 
y (dominance-submission) axes was established, and the continuum of 
these points was assigned the values from -\-56 to —56, with zero at 
the center of the circle (Table 11). Now it was possible to establish 
vertical and horizontal components of each discrepancy. The geo- 
metric distance given by the formula 


(where d^ is the vertical discrepancy and dv the horizontal) is then 
taken as the measure of discrepancy. 

Here another conception of the meaning of discrepancies was con- 
sidered. At all times the discrepancy between any two equally distant 
points should be the same, regardless of the position of the points on 
the circle, i.e., red 1-black 1 should equal red 2-black 2. However, 
from Table 13, the following is noted: for red 1-black 1 the x and y 
discrepancy components are —23 and —5, and for red 2-black 2 they 
are —19 and +13. The squares of each respective discrepancy are 554 
and 530. 


Horizontal (Lev) and Vertical (Dom) Values for Each Octant 

Red I +56 +11 Black 




+ 11 



+ 11 











+ 56 


+ 32 





+ 33 

+ 6 





+ 6 







- 6 



+ 19 


- 6 

+ 33 


+ 19 


Theoretically, the square roots of these numbers should be equal if 
our principle is to be met. The two square roots are 23.537 and 23.022. 
In all such cases encountered, the closest luhole number to both square 
roots was taken to be the discrepancy (in this case, 23). In this way 
fourteen possible discrepancies are obtained. They are presented in 
Table 14. 

All Possible Discrlpancies Around the Pair 1-7 and Their Magnitudes 



1 1 (/ /) 


1 / (i 1) 


7 2 


1 2 (7 2) 


1 2 


7 3 


1 4 


1 5 (7 3) 


7 5 


1 3 


1 4 (7 4) 


1 5 (7 5) 


1 4 


1 5 


Key: The italic numbers refer to moderate (black) diagnostic intensities and the 
ronian-face numerals lo exireme (red) intensities 

The model was then examined in terms of meaning for these dis- 
crepancies and the conceptual-numerical relationships. Table 14 shows 
one set of discrepancies and their jelative magnitudes. 

Careful examination of this table shows the inner relationships of all 
possible distances from one octant. All other octants show the same 
relationships because of the equivalent-distant principle used in estab- 
lishing discrepancy magnitudes. Considering the extreme combina- 
tions only, we have the following set of discrepancies: 1 1 = 00, 


1 2 = 44, 1 3 = 81, 1 4 = 105, I 5 = 114. The same grouping for 
lesser intensity combinations yields the following: 1 1 = 00, 1 2 = 26, 
13^ 48, 1 4 = 62, 1 S ^= 68. Pairwise comparison of equal octant but 
different intensity groups shows that as the distance around the circle 
becomes greater the ratio of the two discrepancies becomes less, i.e., 
00:00, 26:44, 48:81, 62:105, 68:114. This relationship leads to the 
fact that the discrepancy between 1 5 (most extreme of the lesser in- 
tensities) is between the discrepancy for 13 and 12, placing a greater 
discrepancy value for a moderately distant discrepancy of extreme in- 
tensity, i.e., 13, than on- an extremely distant discrepancy of moderate 
intensity, i.e., 1 5. 

Placing all discrepancies in intervals of 20, we have the results in 
Table 15. 


Illustration of the Grouping of All Possible DiscREPANaES 
Involving the Diagnostic Codes 1 and 1 

Code Discrepancy Numerical Value 

1 I a 1) 00-20 

1 i (i 1), 1 2 21-40 

1 2 (i 2) 1 2, / 5 41H50 

1 4, \ 3 {1 I), 1 5 61-80 

I 3, I -^ (/ 4), 1 5 (i 5) 81-100 

1 4, 1 5 101-120 

This grouping shows approximate equations of the various mixed 
discrepancies, such 2.s 1 4 approximately equals 1 3 approximately 
equals 1 5, i.e., the two most extreme moderate-intensity discrepancies 
are approximately equal to the discrepancy between moderate and 
intense of medium distance around the circle. 

A table of weighted scores for each possible interlevel discrepancy 
is presented in Appendix 5. 

This consideration of discrepancy relationships seems to indicate 
that the model we constructed is consistent with the meaning of 
change in terms of the theory of the interpersonal circle. 


Interpersonal Diagnosis of Personality 


Theory of Multilevel Diagnosis 

The preceding thirteen chapters have presented a theory and an em- 
pirical system of personaHty. A compHcated array of variables have 
been described, and the relationships among variables have been classi- 

In this fourth section of the book we are going to apply this system 
of personality to the task of clinical diagnosis and prognosis. The 
numerical code diagnosis, it will be recalled, provides 65,536 personal- 
ity types — at four layers of personality. It is clearly impossible to ex- 
pect to locate sample cases illustrating each of these 65,51)6 multilevel 
combinations. It is equally out of the question to give a clinical de- 
scription of each of these types. 

To use this diagnostic system in clinical situations it is necessary 
only to employ the notion of multilevel analysis and to apply a com- 
mon-sense interpretation of the numerical diagnostic formula. The 
eight-digit diagnostic code, it will be recalled, is nothing more than a 
shorthand summary of the way in which the patient responded at the 
several levels of personality. 

The system is quite complex in the sense that it provides for a great 
variety of types. But the processing of the data and the derivation of 
the diagnostic code is a straightforward clerical, technical (nonpro- 
fessional) task. The interpretation of the diagnostic code is not a de- 
manding assignment since the diagnosis for every level simply denotes 
which interpersonal behaviors the patient manifested. The chnical 
implications follow quite naturally. We simply ask the questions: 
What does it mean if he says this about himself but acts that way? 
What does it mean if he says this and manifests these underlying be- 

Application of the system is facilitated by some clinical experience 
with it, and by some knowledge of the empirical results obtained in 
normative studies. These will be presented in the eight clinical chap- 
ters to follow. 



There is, however, no high-powered theory which has to be mas- 
tered. There is the one assumption that all interpersonal behavior 
serves to reduce anxiety and to maintain self-esteem. The rest is based 
on behavior. What did the patient do, say, indirectly express? A 
multilevel summary of interpersonal behavior yields considerable addi- 
tional infoi:mation about the rigidity of security operations (kind and 
degree) or about conflicts and ambivalences (kind and degree). 

This information is then used to answer functional questions about 
motivation and treatment. 

The system can be seen as a hierarchical pattern of levels which un- 
fold symmetrically. For research or clinical categorization it is con- 
venient to work from the surface into the indirect or deeper areas of 

Single-Level Diagnosis 

In considering an individual case or a general research problem we 
look first at Level L There are sixteen interpersonal types (eight mod- 
erate and eight intense) at this overt behavioral level. Much of our 
research has taken place at this single level. We have attempted to 
discover what probability indices hold for this level. We discover, for 
example, that patients who are hypernormal (code 8) at Level I re- 
main in psychotherapy only half as long as distrustful (code 4) pa- 
tients; and that ulcer patients do not differ significantly from hyper- 
tensive patients at this level. 

Double-Level Diagnosis 

Adding the Level II material we get a much more complicated two- 
layer pattern. First, it should be noted that there are probabiUty find- 
ings which allow us to predict on the basis of Level II alone. When 
we combine the Level I and II indices, new meaning appears. The 
number of possible types multiplies. There are 256 two-level types 
(16 at Level I X 16 at Level II). 

A double-level diagnosis is useful because it points up conflict or 
discrepancy in the presenting fa9ade. Some patients give a dependent, 
fearful symptomatic picture (Level 1 = 6) and may see themselves as 
independent and self-confident (Level II = 2). The code label "62" 
thus becomes loaded with meaning. It points to an ambivalent moti- 
vation, to a marked misperception by the patient of the eff^ect of his 
symptom. It complicates the clinical predictions we are to make about 
the patient since his symptoms (6) are dependent, and his self-regard 
is the opposite (2), 

The double-level diagnosis ''66" forecasts an entirely diff^erent clini- 
cal course. Here, the docile, fearful overt symptomology is duplicated 


by the conscious self -perception. A two-layer commitment to the 
same interpersonal operations is indicated. 

Triple-Level Diagnosis 

The summary code of the subject's fantasy-hero behavior provides 
the third digit for the diagnostic formula. There are sixteen fantasy- 
hero codes (eight moderate and eight extreme). When these are com- 
bined with the double-level codes, a total of 4,096 diagnostic types ex- 
ists (256 double-level types X 16). 

When we consider the third digit in any diagnostic formula, con- 
siderable empirical information is available. We know, for example, 
that the fantasy-hero score predicts future behavior. Thus, the third 
digit is of clinical interest in itself. It indicates what shifts in con- 
scious self-perception we can anticipate. When combined with the 
first digits it fills out a more meaningful pattern. 

A "773," for example, denotes a patient who is friendly and over- 
conventional at the levels of overt presentation and conscious self- 
description. The third digit, "3," indicates a "preconscious" concern 
with hostility. It suggests that underlying antisocial feeUngs exist be- 
neath a fa9ade of bland normahty and that they will probably appear 
in future behavior at the overt levels. 

The clinical meaning of a "773" is very different from a "777." 
The latter maintains a solid, triple-layer structure of affiliative, over- 
conventionality. Self-satisfaction would probably be high and moti- 
vation for therapy low, since the patient cannot tolerate hostile or un- 
conventional feelings at any of the top three layers. A "773" would be 
handled quite differently, clinically, since a conflict exists between a 
conventional fa9ade and underlying "preconscious" sadistic feelings. 

Four-Level Diagnosis 

The fourth digit in the diagnostic code denotes the themes at- 
tributed to fantasy "others." There is less empirical significance or 
clinical meaning attached to this layer. This level has not been studied 
extensively, and no specific empirical significance attaches to it. This 
layer does suggest how rigid or flexible the subject's range of security 
operations is. If themes which are avoided at the top three layers were 
to appear in the fourth code digit, then we might assume that the 
subject does not completely avoid that area. Consider, for example, 
two patients who present triple-layer structures of solid distrust and 
bitterness (444) . One might have a fourth digit of "4," which would 
indicate a complete commitment to schizoid operations. The second 
patient might present an "8" in his "preconscious-other." The code 
"4448" indicates that some tender, responsible feelings exist and can 


be tolerated, at least at this more indirect level of expression, A com- 
mon-sense hypothesis might be that the latter patient would have a 
slightly less pessimistic prognosis than the patient who could not al- 
low any affiliative behavior at any level. 

The fourth digit is, therefore, included in the diagnostic code but 
is given minimal consideration in the clinical sections to follow. 

When the 16 "preconscious-other" codes are combined with the 
4,096 triple-layer types, a total of 65,536 is obtained. 

Organization of the Interpersonal Typology 

A system of interpersonal diagnosis which involves this many types 
may appear bewildering in its scope. We have stressed, however, that 
the system is fairly simple to apply if the common-sense meaning of 
any particular multilevel combination is kept in mind. First the eight- 
digit formula is derived for a patient. To understand the patient's per- 
sonality organization we simply translate the code digits into diagnos- 
tic terms. The conflicts or rigid duplications existing in the multilevel 
pattern will become apparent. 

The eight diagnostic chapters which follow present the clinical and 
research data now available. These chapters refer to the eight typo- 
logical categories at Levels I and IL The "schizoid" chapter is con- 
cerned with patients who present as "44's" at Levels I and II. In 
each clinical chapter the general findings typical of the pure, uncon- 
flicted case will be presented. 


Adjustment Through Rebellion: 
The Distrustful Personality ' 

This chapter deals with those individuals who select distrust and re- 
bellion as their solutions to life's problems. This is the "44" personal- 
ity type. In their crucial relationships with others, these human beings 
consistently maintain attitudes of resentment and deprivation. They 
handle anxiety by establishing distance between themselves and others. 
At the critical moments of relationship with others they become cyni- 
cal, passively resistant, and bitter. 

The distrustful way of life is in some ways a puzzling phenomenon. 
The ideals of our culture stress adjustment, closeness, and cooperation. 
It is generally taken for granted that trustful, loving relations with cer- 
tain important others is one of the basic human goals. There exists, 
however, a very large group of individuals who consistently avoid 
this relationship. They compulsively eschew closeness with others. 
They are traumatized and threatened by positive feelings. 

These human beings often do not voluntarily seek distance and dis- 
appointment from others. In their conscious ideals, on the contrary, 
they may strive and long for tenderness. They are usually frustrated, 
depressed, and most dissatisfied with their situations. 

They regularly manifest, however, the reflexes of distrust and re- 
sentment. They involuntarily provoke rejection and punishment from 
others. They cannot tolerate durable relationships of conformity or 

The Purpose of Distrustful Behavior 

Those human beings who are overtly bitter and cynical have se- 
lected these operations because they find them most effective in ward- 

* In this chapter and the subsequent seven, we shall be discussing pure interpersonal 
types based on Level I-M and II-C diagnosis. We shall describe the unconflicted sub- 
ject who presents the same security operations in his symptomatic behavior and in his 
conscious self-descriptions. Space does not permit a consideration of the conflicted 



ing off anxiety. Pain and discomfort are traditionally associated with 
alienation from others, but for these subjects this discomfort is less 
than the anxiety involved in trustful, tender feehngs. For the person 
who has experienced past rejections or humiliations there are certain 
comforts and rewards in developing a rebellious protection. The 
essence of this security operation is a malevolent rejection of con- 
ventionality. Trust in others, cooperation, agreeability, and affilia- 
tion seem to involve a certain loss of individuahty. Giving or sharing 
or trusting requires a sacrifice of pure narcissism and some relinquish- 
ing of the critical function. 

The rebellious adjustment provides a feehng of difference and 
uniqueness which is most rewarding to some individuals. Inevitable 
ties and responsibilities go with an agreeable, conventional adjustment. 
For the person who avoids this way of life there are certain rewards 
— a rebellious freedom, a retaliatory pleasure in rejecting the conven- 
tional, a delight in challenging the taboos, commitments, and expecta- 
tions which are generally connected with a durable affiliative rela- 

In the extreme case, the security opeiations of distrustful aliena- 
tion involve a spiteful and bitter rejection of love and closeness. This 
phenomenon has been best understood by Sullivan. He has given a 
most thoughtful description of this process: 

Some years ago, the young nephew of one of my friends was admitted to 
the Henry Phipps Psychiatric Clinic. The patient was suffering an acute 
schizophrenic disturbance, catatonic in type. He was placed under principal 
care of a close friend of mine, and I followed developments closely and saw the 
patient occasionally. As he became unmanageable, he was transferred to the 
Sheppard and Enoch Pratt Hospital, arriving there mute and requiring feeding 
by the nasal tube. He was extremely resistive to this feeding unless I did it, in 
which latter case he came to help with the insertion of the tube. I thoughtlessly 
took over on all these occasions and otherwise greatly interested myself in him. 
As he was convalescing quite nicely, he underwent what I call a malevolent 
transformation of interpersonal relations and became first mischievous and later 
definitely "hateful" on the ward. The outcome was a chronic dilapidating ill- 
ness requiring State Hospital care. 

From the few facts recited above and sundry other observations in my own 
and, mediately, other psychiatrists' work I inferred the theory of malevolent 
transformation of "personality," now taught in the Washington School of Psy- 
chiatry, after considerable supporting evidence as to its current adequacy had 
been derived from data on personality development. 

In brief, this theory holds that if one progresses into a relatively enduring 
situation in which one's indicated needs for tenderness are customarily re- 
buffed, one comes to manifest malevolent behavior when one needs tenderness, 
in lieu of showing the need, and to expect— and by this pattern all but guarantee 
—an unfavorable attitude towards one in others. (6, pp. 451-52) 


The purpose of the malevolent transformation, we assume, is to 
avoid the intense anxiety created by the patient's tender feeHngs. 
These patients apparently have come to expect that loving feelings in 
themselves or in others are the prelude to anxiety and rejection. The 
reflexes of bitter distrust resolve this dilemma very nicely. Such re- 
flexes w^ard oflF one's own trustful feelings and tend to push away the 
other person. 

In moderate intensity the "44" security operations of rebellious 
skepticism have certain adaptive advantages for the individual and for 
society. They are associated with a healthy, critical approach to the 
accepted conventions and to the accepted forms of social relationship. 
There is a familiar observation that every creative expression is an act 
of rebellion, a critical questioning of some conventional concept. 
Skepticism gives the human being a sense of freedom and uniqueness. 
It protects against surprises. A mildly disappointed cynicism is an ex- 
cellent preparation for future disappointments. 

The critical, rebellious person can play a most healthy role in any 
social group. Docile inertia or fearful-need-to-conform or need-to-be- 
liked can lead to a stultifying atmosphere. There are valuable rewards 
for the successful rebel who maintains a realistic, accurate skepticism 
toward the accepted ways of doing things. 

James Joyce has provided an interesting illustration of this rela- 
tionship between bitter rebellion and creativity. When his hero dedi- 
cates his Hfe to art he adopts the motto non credo, non serviam and 
recognizes that this rejection of family, church, and society commits 
him to a life of "silence, exile and cunning." 

The "44" mode of adjustment has been eulogized by many writers. 
Its most enthusiastic advocate is Robert Lindner. He states: "It is pos- 
sible, then, to escape from history, to break out of the cage whose 
outer limits never have worn smooth and deeply grooved with endless 
pacing. And it is possible to do this without the letting of blood, with- 
out violence, without the sacrifice of basic values. All that is re- 
quired is to reach for one cup wherein the heady mixture of true re- 
bellion, the brew of sweet life-affirming protest, has been poured, 
for this — and this alone — is the elixir vitae." (3, p. 296) 

This author has taken one mode of adjustment (at one level of per- 
sonality) and has made it the key to mental health. In the Kaiser 
Foundation system, the overt reflex security operation of rebellious 
nonconformity is one of eight generic security operations, each of 
which has an adaptive and a maladaptive intensity. 

Skeptical alienation from convention and from acceptance of others 
can serve several purposes for the individual who selects this way of 


life. These include: protection for disappointment, realistic critical 
rejection of the conventional, the warding ojff of anxiety generated by 
trust and tenderness, the freedom associated with uniqueness and re- 
bellious individuality, and, in the pathological extreme, malevolent re- 
taliation for the feelings of rejection by society in general or specific 
"other ones." 

The Effect of Distrustful Behavior 

Bitter rebellious behavior pulls punitive rejection and superiority 
from others. In systematic language, FG provokes BCD; crime pro- 
vokes, punishment. 

In the passage just quoted Sullivan has described this phenomenon 
very clearly. He speaks of this pattern almost guaranteeing an un- 
favorable attitude in others. A sour, distrustful approach invariably 
establishes distance from others, provoking them to ignore, condemn, 
or disaffiliate. 

In the case of the adaptively, moderately rebellious person the same 
reaction develops to a milder degree. We consider here the individual 
who communicates in his actions, his demeanor, and his interpersonal 
reflexes a message of skepticism and passive rejection of conventional- 
ity. These persons are seen as iconoclastic, eccentric, different, creat- 
ive. Originality is inevitably linked to rebellion, i.e., rejection of the 
established, the authoritative, the conventional. The iconoclastic ap- 
proach usually pulls irritated rejection from those who represent au- 
thority and from those who conform to it. 

One of the most consistent and interesting results of the Kaiser 
Foundation research has been the empirical importance attached to 
the conformity-nonconformity axis of the interpersonal diagnostic 
circle. Conventionality (as measured by the points L, M, and N on the 
circle at Levels I and II) is closely related to absence of overt anxiety, 
to the presence of psychosomatic symptoms, to a state of low moti- 
vation for psychotherapy, and to many other personality variables 
(see Chapter 18). The nonconventional operations of distrust, rebel- 
lion, and alienation are defined by the opposite end of the LMN axis, 
i.e., by the points F and G on the circle. 

The individuals whose overt operations emphasize nonconformity 
and skeptical distrust invariably isolate and alienate themselves from 
others. Conventional people are often irritated and made anxious by 
the sullen, rebel. Even the most agreeable and overtly friendly souls 
can be provoked to disapproval when faced with distrustful opera- 

The psychotherapy group provides an excellent locale for observ- 


ing these processes. Group members are quickly trained to reject or 
isolate themselves from the sullen patient. By their tone of voice, their 
gestures, often by their dress, these patients communicate the message, 
"I am different; I distrust and disagree with you." 

The principle of reciprocal relations operates in the case of the re- 
bellious personality with impressive and depressing results. These pa- 
tients provoke disregard and hostility from others. This behavior on 
the part of others leads to an increase in retaliatory distrust. The sul- 
len, distrustful person creates for himself a world of punitive rejection. 

These reciprocal processes do not work with uniform consistency. 
There are some individuals who are so committed to friendly, nur- 
turant responses that they do not immediately react with hostility 
when faced with distrustful reflexes in another. They may attempt to 
win the sullen person over into a close relationship. Where the rebel- 
lious fagade is adaptable and not extreme, this may lead to a relaxation 
of the distrustful defenses. This often happens in social and thera- 
peutic experiences. 

Where the distrustful reflexes are intense and are the sole means of 
warding off anxiety, then positive feelings in the "other one" tend to 
be rebuffed. This bitter reaction will eventually discourage the most 
persistently friendly "other" and will inevitably lead to irritation. 

The severely distrustful person is most comfortable when he is ex- 
pressing bitter feelings. He is threatened and suspicious of tenderness 
which can be viewed as an intolerable threat to his mode of adjust- 
ment. The common assumption that what the deprived, distrustful 
person needs is love and affection can be seen to be a well-meaning but 
naive notion. To the person with a set of severely crippled reflexes 
tenderness in the "other one" is a loaded gun — a most frightening and 
fearful stimulus. The "malevolent transformation" described by Sulli- 
van is often the reaction to the threat of affection. 

D. H. Lawrence has provided us with a clear illustration of the way 
in which the distrustful, disaffiliated person avoids tender feelings. 
The hero of Aaron's Rod announces: "I don't want my Fate or my 
Providence to treat me well. I don't want kindness or love. I don't 
believe in harmony and people loving one another. I believe in the 
fight and in nothing else. I believe in the fight which is in everything. 
And if it is a question of women, I believe in the fight of love, even if it 
blinds me. And if it is a question of the world, I believe in fighting it 
and in having it hate me, even if it breaks my legs. I want the world to 
hate me, because I can't bear the thought that it might love me. For 
of all things love is the most deadly to me, and especially from such a 
repulsive world as I think this is. . . ." (1, pp. 307-8) 


Clinical Manifestation of Distrust and Rebellion ^ 

The symptomatic correlates of this mode of overt adjustment are 
quite typical, and clearly different from other diagnostic types. 

These patients exhibit sour, pessimistic, or indifferent feelings. This 
may often appear to be a fiat affect or an absence of feelings. This is 
probably an incomplete and misguided interpretation. There is no 
evidence to indicate that the distrustful person feels less intensely. It 
is necessary to look at the interpersonal implications of a resigned or 
skeptical approach. These patients do not admit to conventional reac- 
tions. Their nonconformist facade means that they express different 
feelings in different ways. They are communicating by their actions 
and their verbalizations an intense and emotionally loaded message of 
sullen distrust. 

Clinically this attitude may be expressed in the generic motto: "I am 
a sullen, disappointed person; you can't do anything for me." 

These patients do not participate in therapeutic planning with 
docile eagerness or enthusiastic hope. They may agree to treatment, 
but the note of skeptical passive resistance is often obvious. 

In regard to symptoms, these patients tend not to have psychoso- 
matic ailments;^ nor do they complain of the overt anxiety of the 
phobic or the worries of the obsessive. They present characterological 
or straightforward interpersonal disorders. They tend to complain of 
marital discord, social isolation, frustration, distance and disappoint- 
ment in their relations with others. A most typical symptom is occu- 
pational or academic difficulty. They may describe a history of re- 
bellion against authority, and are often stalemated in their vocation. 
They are frank to admit their disillusionment and irritation with others. 
They tend to complain of their treatment at the hands of others; yet, 
in contrast to some of the poignant masochists described in the foUow- 

^ In this secdon and in the "Clinical Manifestation" sections of the following seven 
chapters, we shall consider the symptomatic pictures presented by the various diag- 
nostic types. These discussions are highly generalized and suggestive. Two qualifica- 
tions must be kept in mind. First, we are considering here the symptomatic picture 
of the pure type (in this chapter the "44"). Variations in behavior at other levels can 
change the symptomatic presentation; thus, the "41" comes to the clinic presenting a 
facade diflferent from the "44." The second qualification refers to the precipitating 
cause for psychiatric referral. Most of the patients coming to the psychiatric clinic 
are in some state of anxiety. Often something has happened recently to threaten their 
overt security operations (whether they are schizoid or hysterical). We are con- 
sidering, in this section, the general clinical impression made by the patient which is 
often quite different from the "current" anxiety which brings him to the clinic. 

^ In one diagnostic study comparing the Level I interpersonal diagnoses of a group 
of psychosomaoc and neurotic patients, only 7 per cent of the psychosomatics fell into 
the rebellious-distrustful octant {FC) of the diagnostic grid, whereas 43 per cent fell 
into the opposite sector. (2) 


ing chapter, they do not attempt to win pity or to present them- 
selves as good and blameless. They stress instead a grievance against 
the M^orld, a pessimistic disappointment with self and others. 

These security operations, it will be noted, do not lend themselves 
to a well-motivated, eager acceptance of psychotherapy. They often 
agree to treatment with a half-hearted pessimism: "I guess I'll have to; 
I don't see any other solution," etc. These patients often express pas- 
sive complaints about the kind of therapy offered, about the therapist 
to whom they are assigned, about the necessity to be in a clinic, etc. 
Often these patients will sullenly refuse the therapy that is recom- 
mended. For example, they may interpret the assignment to group 
therapy as a sign of rejection by the clinic. 

In the case of the moderate rebel, these gloomy, resistant operations 
may not become apparent. They may employ a sarcastic, self-immo- 
lating humor. They may describe their isolation and disappointment 
with a bitter, wry irony. If they sense honesty and reasonability in 
the clinician they may muffle or shelve their skepticism. 

Regardless of the intensity or rigidity of the character structure, 
there is one interpersonal rule which invariably holds for the "44" 
personality. They are painfully sensitive to phoniness, pomposity, 
naive obtuseness, or arrogance on the part of the "other one." These 
patients tend, as a group, to load their perceptions of others with a 
hostile skepticism. They look for dishonesty and hostility in others. 
They are incredibly sensitive instruments for picking up rejection or 
punitive feelings in others. Naive hysterical patients, on the contrary, 
tend to act on the assumption that others (in their in-group) are con- 
ventional and sweet like themselves. 

We have noted in Chapter 7 that all maladjusted persons are skilled 
in provoking others to certain reciprocal responses. The distrustful 
patient is most accomplished in pulling bureaucratic or moral disap- 
proval from others. He often puts the therapist to elaborate tests 
aimed at provoking impatience or moral censure. He compulsively 
clings to the often automatic and involuntary conviction that the 
clinician fails to understand him, or acts in a pompous, overconven- 
tional manner. He specializes in provoking the therapist to set limits 
and re-create an authority-rebellion or rejecting-distrustful relation- 

The distrustful "44" personality type described in this chapter has 
certain similarities to a behavior pattern observed in group psycho- 
therapy by Jerome Frank et al. (5, pp. 215) Frank calls this type the 
"help-rejecting complainer" and states that the pattern "consists of a 
patient's continuing attempt in the group to elicit help — often without 
actually asking for it — and his attempt to prove greater need than 


other people, while either imphcitly or explicitly rejecting all help 
offered. This pattern seems to be an expression of conflict between 
the patient's perception of himself as needing help and his anger at 
all potential help-givers for being unable or unwilling to supply it. 
His behavior justifies his anger toward the help-givers and maintains 
his claim for help while preventing him from becoming dependent 
on the distrusted potential help-givers." 

We have so far stressed the symptomatic and interpersonal aspects 
of the clinical picture. There are certain psychometric correlates of 
the rebellious presentation which appear on personality tests — for 
example, the MMPI and the Rorschach — which are independent of 
the interpersonal system. 

Patients who behave in a sullen, distrustful manner (Level I) have 
a typical pattern on the MMPL Their high peaks fall on depression, 
schizoid, and psychopathic scales. They also have elevations on the F 
scale which is a rough measure of nonconformity. They generally 
do not have elevations on the L, K, Hy, and Hs scales. 

This suggests that pessimistic dysphoria (D), alienation (Sc), rebel- 
lious disidentification (Pd), and nonconventionality (F) are character- 
istics of the distrustful personality. The scales on which they show 
low scores are those related to denial of antisocial or hostile tendencies 
and to a naive, conventional, sweet fagade. The distrustful personality 
can be differentiated on the MMPI from the obsessive-masochist. The 
latter have pronounced depression and psychasthenic scales. The 
former exhibit schizoid scores which are higher than psychasthenia; 
and, though the depression scores are elevated, they are not as marked. 
The higher the F, the more likely that rebellion and not masochism is 
the security operation. 

On the Rorschach or TAT these patients characteristically manifest 
different, odd, idiosyncratic content. Unconventional themes are 
common — bizarre situations, freely described sexual themes, and poor 
form responses. 

Interpersonal Definition of the Schizoid Maladjustment 

Chapter 12 presented evidence that certain standard psychiatric 
diagnoses were related to specific interpersonal patterns. Extreme, 
imbalanced social patterns thus can help to establish psychiatric diag- 

Distrustful, intensely rebellious behavior is characteristic of the 
schizoid personality. Such a personality shows maladjustment essen- 
tially in bitter, disappointed alienation and tends to handle anxiety by 
avoiding close, tender contacts with other individuals and by avoiding 
close commitments to society in general. In the extreme case this be- 


comes a malevolent rejection of people and of conventional social 

Many of the symptoms of the schizoid condition may be inter- 
preted in the light of interpersonal communication. They seem to be 
expressions of a bitter alienation from accepted standards, a refusal to 
conform which in the extreme case becomes a rigid pathological in- 
ability to conform. Highly individualistic, eccentric behavior is (in 
the absence of organic disease) generally pathognomonic of schizo- 
phrenia. It seems possible ro think of this as being an intense, and often 
desperate, attempt to express difference, to establish a complete, bit- 
ter break from conventional reality. 

Kobler, speaking from therapeutic experience with schizophrenics 
in the Pinel Foundation Hospital, believes that schizoid malevolence 
can be seen as asking the question, "Even if I do this and am Uke this, 
can you still love me?" The therapeutic staff at Pinel senses hope and 
a violent testing of the therapist with the anticipation of, "No, the 
other cannot love." When the reaction of the therapist is not rejec- 
tion, Kobler states that there is further negativistic testing and at the 
same time the continued hope of finally finding the one who will not 

The Schizoid Psychosis 

The interpersonal effect of bizarre behavior is to provoke exas- 
perated rejection from others. Marked eccentricity flaunts to the 
world the message, "I do not accept your ideals of conduct; I do not 
conform. I do not want your approval." This usually guarantees to 
the subject the disapproval of others. In the extreme case (psychosis) 
it provokes society to punitive incarceration. 

The tendency for bizarre behavior to pull rejection from others 
was illustrated by the reactions of some therapy group members to the 
schizoid fantasies of a fellow member. This particular group had 
been meeting for almost a year and an unusually frank, honest recog- 
nition and acceptance of each other had been developed. One of the 
members was a chronic, severe, ex-state-hospital schizoid who had ap- 
parently never been able to integrate a friendly, trusting relationship 
with another human being. The group had initially ignored and de- 
spised her. By the fourth month, her ability to train others to reject 
her was the focus of considerable study. Even after this had been 
worked through for two more months, she could be reduced to panic 
by a warm smile or casual friendly compliment. 

In one session a woman who employed narcissistic, exhibitionistic 
operations described her fantasy of parading down the street in glam- 
orous clothes in order to provoke envy and admiration from others. 


The schizoid woman then confided the repeated fantasy of run- 
ning naked out into the street. The question was then posed to the 
group what reaction would be provoked from them at the sight of a 
neighbor running naked in the street. Their associations were, "I'd 
think she's nuts," "I'd call the cops to come and take her away," etc. 
These associations demonstrated the effect of the bizarre fantasy in 
pulling rejection and intolerant contempt from others. Expressed in 
the context of an honest and accepting group they sharpened the 
schizoid patient's understanding of the rebellious and alienating effect 
of her security operations. 

Many schizoid or schizophrenic symptoms can be interpreted as 
interpersonal communications conveying to others the theme of ma- 
levolent disaffiliation. The inability or refusal to integrate close rela- 
tions with others and the tendency to perceive and react differently 
have such a consistent and inevitable impact on others that they seem 
to confirm the hypothesis that an interpersonal purpose is involved. 
In the extreme case these desperate violations of customs and accepted 
social patterns become the symptoms of psychotic negativism; autism, 
incontinence, refusal to eat, etc. The bitter, rebellious anger involved 
in these behaviors has often been commented on by clinicians. 

Some interesting complications are introduced by cultural differ- 
ences. A schizophrenic psychosis is defined as a desperate, repetitious, 
malevolent, distrustful rebellion. Now, the behaviors which express 
these motives may differ from one society to another. Thus, failure to 
eat and an insistence upon the reality of one's own fantasy life are, in 
our society, negativistic and alienating behaviors. In another society 
the same behavior may be symptomatic of an extreme desperate at- 
tempt to overconform. It may express the message of frantic re- 
ligious overconventionality. The interpersonal effect of the symptom 
is the key to its diagnostic meaning. 

A glance at the symptomatic signs of schizophrenia listed in any 
psychiatric text seems to suggest that most of them are calculated to 
provoke frustration and irritation in others. The symptoms of the 
obsessive state, by comparison, tend to provoke feelings of superiority 
in the other one. 

The rebellious implication of the schizoid maladjustment has been 
noted by other writers. Powdermaker for example writes: 

Why does the schizophrenic use the particular defenses that he does against 
these fears and conflicts? Why does he make himself ununderstandable and 
so different from the social norm in his relationships, instead of endeavoring to 
conform to the social norm as the neurotic does? That the schizophrenic is an 
unsuccessful rebel appears to be one of the outstanding aspects of his behavior. 
This was pointed out in the work of Ackerly, in which he showed how the dc- 


linquent acts of some of his adolescent patients had s^ved them from a probable 
schizophrenic breakdown. (4, pp. 61-62) 

Research Findings Characteristic of the Distrustful Personality 

Here is a summary of some of the studies accomplished on the 
schizoid personality by the Kaiser Foundation project. 

1. Patients who exhibit rebellious distrust in their overt operations 
do not tend to have psychosomatic symptoms. 

2. Psychosomatic patients do not tend to utilize these interpersonal 
operations at Levels I or IL 

3. Distrust at Levels I and II is related to depression (D), noncon- 
formity (F), schizoid tendencies (Sc), and rebellious disidentification 
(Pd) on the MA4PI. 

4. These patients are among the initially best motivated for psycho- 
therapy. They do not tend to terminate their clinic contacts after 
evaluation but are likely to go on into treatment and to remain in treat- 

5. They are (along with the psychopathic personalities) the most 
consciously disidentified with their mothers and their fathers. 

6. They tend as a group to be extremely disidentified with their 

7. They (along with the psychopathic personalities) show a ten- 
dency to misperceive the interpersonal behavior of others. They are 
inclined to attribute too much hostility to others. 

8. Considering all the eight diagnostic types (at Level I), the 
schizoid group comprises the largest number of unmarried individuals. 
This suggests that more schizoid patients than patients of any other 
diagnostic type have failed to accomplish a durable, conventional mat- 
ing relationship. 

9. The schizoid personality tends to appear in certain cultural and 
institutional samples much more frequently than others. The per- 
centage of rebellious individuals (Level I-M) in various samples is 
presented in Table 16. The percentage figure expected by chance for 
these groups is 12.5. It will be noted that eight groups contain more 
than or close to the expected number of rebellious-distrustful person- 
alities — the four psychiatric samples, the graduate student, the overtly 
neurotic dermatitis, the prisoner, and the psychotic samples. All but 
two of these define "people in trouble," i.e., at odds with or malad- 
justed to society. The fact that the graduate student sample contains 
a higher percentage of schizoid personalities than the more conven- 
tional groups suggests that rebelliousness is a characteristic of this sam- 
ple. This is an interesting confirmation of the hypothesis that creativ- 
ity, delinquency, and alienation involve somewhat similar security 



Percentage of Rebellious-Distrustful Personalities (Level I-M) 

Found in Several Cultural Samples 

% of Rebellious-Distrustful 

Institutional or Symptomatic Sample 



Psychiatric Clinic Admission 



CoUege Undergraduates 



University Psychiatric Clinic 



Middle Class Obese Patients (Female) 



Overtly Neurotic Dermatitis Patients 



Self-inflicted Dermatitis Patients 



Unanxious Dermatius Patients 



Group Psychotherapy Patients 



Individual Psychotherapy Patients 



Hypertensive Patients 


Ulcer Patients 


Medical Control Patients 



University Counseling Center 



University Graduate Students (Male) 



Stockade Prisoners (Male) 



Hospitalized Psychotic Patients 



Officers in Military Service 





Operations — rebelliousness toward conventionality. The overtly neu- 
rotic dermatitis sample differs from the other psychosomatic groups 
in the amount of bitter, masochistic behavior manifested. Rebellious- 
distrustful patients comprise the largest percentage of patients who 
enter and remain in group therapy at the Kaiser Foundation clinic. 
One out of every four patients (24 per cent) entering group therapy 
were schizoid personalities. Two factors are suggested to account for 
this finding. The clinic intake conference has found that group 
therapy is the most effective therapeutic agent for patients who are 
isolated, distrustful, and distant from others. Thus, the clinicians re- 
fer more schizoid patients to groups. Obsessives and phobic patients, 
on the contrary, are more likely to be referred to individual treatment. 

A second possible explanation for this finding is that schizoid per- 
sonalities tend to remain in therapy because the diffused and diluted 
transference phenomena in the group are less intense than the trans- 
ference of individual therapy. These distrustful people can apparently 
stand the interpersonal pressure in the groups where they can remain 
silent or sullen for considerable periods without completely disrupt- 
ing the therapeutic process. 

10. The percentage of subjects who diagnose themselves as re- 
bellious-distrustful (Level II-C) is presented in Table 17. These find- 
ings tend to be in line with the Level I-M data just discussed. Group 

% of Rebellious-Distrustful 




















Percentage of Rebellious-Distrustful Personalities (Level II-C) 
Found in Several Cultural Samples 

Institutional or Symptomatic Savrple 
Psychiatric Clinic Admissions 
Hospitalized Psychotic Patients (Male) 
Group Psychotherapy Patients 
Individual Psychotherapy Panents 

Overtly Neurotic Dermatitis Patients 
Self-inflicted Dermatitis Patients 
Unanxious Dermatitis Patients 
Medical Control Patients 

Ulcer Patients 

Hypertensive Patients 

Middle Class Obese Patients (Female) 

Total 781 

psychotherapy patients again comprise the largest percentage of 
schizoid cases. Overtly neurotic dermatitis patients, again, differ from 
the other psychosomatic groups in the emphasis on bitter behavior. 

1 1 . Schizoid patients consciously perceive their parents to be v/t2k 
and distrustful people. The mean placement of fathers of schizoid 
patients is in the FG section of the diagnostic grid. Mothers locate 
in the guilty, self-punitive {HI) sector. Schizoid patients report their 
marital partners as strong and exploitive. 


1. Lawrence, D H. Aaron^s Rod. New York- Thomas Seltzer, 1922. 

2. Leary, T., and H. Coffey. Interpersonal diagnosis: Some problems of methodol- 
ogy and validauon. /. abnorm. soc. Psychol., 1955, SO, No. 1, 110-25. 

3. Lindner, R. Prescription for rebellion. New York: Rinehart Press, 1952. 

4. Powdermaker, Florence. Concepts found useful in treatment of schizoid and 
ambulatory schizophrenic patients. Psychiat., 1952, 25, No. 1, 61-71. 

5. Rosenthal, D., J. Frank and E. Nash. The self-righteous moralist in early meet- 
ings of therapeutic groups. Psychiat., 1954, 11, No. 3, 215-23. 

6. Sullivan, H. S. Therapeutic investigations on schizophrenia. In P. Mullahy 
(ed.), A study of interperso7ial relations. New York: Hermitage House, 1949. 
Copyright, New York: Thomas Nelson & Sons. 


Adjustment Through Self-Effacement: 
The Masochistic Personality 

We are considering in this chapter the many personality types which, 
despite their multilevel differences, have one important thing in com- 
mon — they all present in their overt operations a fagade of self-efface- 
ment. This is the ''55'' personality type. 

The message which they communicate to others in their face-to- 
face relations is "I am a weak, inferior person." Through their auto- 
matic reflex operations they train others to look down upon them 
with varying intensities of derogation and superiority. 

The mild form of this security operation is manifested as a modest, 
unpretentious reserve. In its maladaptive extremes it becomes a maso- 
chistic self-abasement. In either case the person employing this gen- 
eral mechanism avoids anxiety by means of retiring, embarrassed 
diffidence. He is automatically mobilized to shun the appearance of 
outward strength and pride. 

The Purpose of Self-Depreciation 

The individuals who employ this security operation do so because 
they feel that this social role is the safest and least dangerous position 
to be assumed in this particular situation. Now persons vary in the 
consistency with which they employ any interpersonal behavior. 
Some repetitiously respond with the same reflexes in almost all situa- 
tions, whether appropriate or not. Others may automatically assume 
modest, retiring reflexes in particular situations where they expect it 
to be appropriate. Many subjects, for example, act embarrassed and 
reserved when facing strong and potentially dangerous others. 

In this chapter we are considering those patients who present a 
fagade of guilty submissiveness in their approach to the clinic. We 
cannot assume, of course, that all these patients act in this wav in all 


their life relationships. We simply know that this is their inter- 
personal impact on the clinic. It is, therefore, the aspect of their per- 
sonality that we must begin to respond to and deal with. 

Whenever we observe or measure this security operation, we may 
assume that an individual has learned to employ self-depreciation as a 
protective device in certain situations, or in all situations. Later in- 
vestigation (e.g., measurements at other levels) will indicate the range 
and consistency of this security operation. 

The role of masochism in contributing to the security of the indi- 
viduals has been pointed out by several psychoanalytic authors. 
Menaker has contributed an excellent summary of these theories: 

The observation that masochism is a way of avoiding anxiety, a point on 
which a number of analysts agree, is a clue to the fact that one of its important 
aspects is its function of defending the ego. Important psychoanalytic contribu- 
tions to the understanding of masochism, however, have thus far been too ex- 
clusively concerned with its libidinal meaning. The point of departure has 
been how gratification is achieved for the individual through masochistic be- 
havior, rather than examining the way in which it serves the ego. 

We find that viewing the problem of masochism from the standpoint of the 
self-preservative functions of the ego leads to new insights. As might be ex- 
pected, the ego function of the masochistic attitude is most clearly discernible in 
the study of moral masochism. Berliner, confining his observations primarily 
to moral masochism, has made an important contribution to the concept of 
masochism as a defense mechanism of the ego. He takes masochism out of the 
sphere of the instincts and views it as a function of the ego. It is 'a pathologic 
way of loving' in which the ego through processes of introjection, identifica- 
tion and superego formation turns the sadism of the love object (not its own 
sadism) on itself. The motivation for so doing is the need to cling to a vitally 
needed love object. The dependent child accepts the suffering emanating from 
the rejecting love object as if it were love, failing to be conscious of, or denying 
the difference between, love and hate. Once the hating love object has become 
part of the superego, the constant wish to please and placate the superego causes 
the individual to lose his identity and to 'make himself as unlovable as he feels 
the parent wants him to be.' 

Analytic experience confirms Berliner's view of masochism as a function of 
the ego in the service of maintaining a vitally needed love relationship to a 
primary object. (4 pp. 207-8) 

The general purpose of the masochistic mechanism seems to involve 
the warding off of anxiety by means of self-depreciation. The more 
specific meanings of the mechanism vary from case to case depending 
on the multilevel pattern. 

The fact that the rather shallow methodology of the interpersonal 
system defines several thousand types which express masochism at one 
or more levels of personality testifies to the difficulty of making broad 
generalizations about the specific meaning of masochism. The pattern 



of conscious and "preconscious" identifications give different inteqjre- 
tations of self-punitive behavior. The introjection patterns are also 
crucial in some cases. We have developed one hypothesis which is in 
line with the psychoanalytic theories summarized above. It seems 
logical to assume that wherever masochism is expressed at any level 
of personality, then sadistic feelings are also present. These may be 
attributed to the conscious or "preconscious" perceptions of others or 
they may be restricted to deeper levels of "self-behavior." Guilt does 
not exist without some introjection or underlying acceptance of puni- 
tive themes. Self-criticism seems inevitably to involve some aspect of 
hostile criticism expressed against or projected on others. 

It must be kept clear that we are discussing involuntary reflexes at 
this point. We are not referring to the conscious, deliberate assuming 
of a humble role — nor to the expression of modest words (i.e.. Level 
II humility) ; we are thinking rather of automatic tendencies to handle 
insecurity by means of weak, depressive, shy operations. 

The Effect of the "'55'' Security Operations 

Self-effacement pulls depreciation and patronizing superiority from 
others. (In the code-language of the interpersonal system, HI pulls 
BC and DE from others.) That is to say, if a person acts in a glum, 
guilty, withdrawn, and weak manner, he will tend to train others to 
look down on him and to view him with varying amounts of contempt. 

One interesting expression of masochistic behavior which invariably 
provokes others to scorn is the "buffoon" personality. One psycho- 
analytic interpretation of the interpersonal meaning of the clown's 
behavior points to the assumption of the castrated role. According to 
Grotjohn (1) the clown in his dress, gestures, and thematic expres- 
sions is telling the audience: "I am a harmless, weak, defeated person." 
The social buffoon seems to exhibit his shameful, inferior position and 
to force the onlookers to laugh at him and to patronize him. 

The reciprocal interaction does not occur in every case. The 
phenomenon of reciprocity is, as we have seen in Chapter 7, a prob- 
ability statement. Self-derogation sometimes pulls initial sympathy, 
but if the guilty reflex does not shift in response to this positive re- 
action, the "other one" will inevitably respond with irritation and dis- 
approval. Another factor preventing the reciprocal process from in- 
variably working resides in the personality of the "other one." If a 
modest person, or a buffoon, is dealing with a rigidly docile "other" 
— the latter may not respond with superiority and disdain. In general 
these relationships do not remain durable since the self-depreciator 
tends to gravitate away from "equal" relationships and to prove re- 
jection by means of withdrawal. The docile person tends also to avoid 


equal relationships and to seek strong, guiding partners. If two indi- 
viduals with submissive fagades maintain a durable relationship, it will 
generally be found that a reciprocity of underlying themes (often 
of a competitive or depreciatory nature) is preserving the interaction. 

Self-abasing individuals provoke punitive and arrogantly superior 
reactions from others. Most persons do not prefer to maintain rela- 
tionships with weak, guilty people. They tend to look down on the 
masochists when they encounter them and do not enter into durable 

While most people avoid the masochists, there are, however, some 
dramatic exceptions to this rule. By the systematic and statistical logic 
of the interpersonal circle, one quarter of the population is bound to 
fall into the upper left-hand quadrant. This is the area which includes 
the operations of exploitation (C), narcissism (B), and punitive hos- 
tility (D). These individuals provoke fear, envy, and guilt in others. 
The modest masochist, we have seen, trains others to reject and despise 
him. The beautiful interlocking of reciprocal reflexes which occurs in 
the relationship between these two types is, of course, one of the 
most familiar problems in dynamic psychology. 

Self-effacing, guilty individuals feel the least anxiety when they are 
manifesting their masochistic reflexes. They therefore gravitate to 
and stay with those individuals who will provoke the least anxiety — 
the aggressive, exploitive characters from the upper left part of the 
diagnostic grid. 

This reciprocal phenomenon is seen over and over again in the mal- 
adaptive masochistic marriage — the overtly sorrowful, martyred, 
abasive wife hopelessly entangled with the brutal husband — or the 
overtly shy, timid man wearing himself out in service of the exploitive, 
narcissistic wife. 

This exchange of guilt and superiority also exists with remarkable 
frequency in the relationships of normal, adaptive individuals. One 
individual takes the modest, inferior, self-eflFacing part, while the 
other exercises the superior role — to the comfort of both. Such 
reciprocal relations are generally complicated by underlying motives. 
We are discussing in this section the general aspects of the modest 
masochistic security technique as exhibited in overt interpersonal re- 
flexes. We shall therefore postpone the detailed discussion of the 
multilevel patterns which usually underly the self-abasive fa9ade. 

Level I modesty and self-depreciation can be adaptive or rigid, mal- 
adaptive responses. Their purpose is to ward off anxiety. They lead 
to the counterreactions of depreciation and superiority on the part of 
the "other one." Extreme, rigid masochism invariably sets up new 
chains of conflict and increased anxiety which can be responded to by 


increased repetition of self-abasement, by related symptomatology, 
and by other signs of psychic distress. 

We shall now consider some of the clinical manifestations, both 
interpersonal and symptomatological, of the modest-masochistic per- 

Clinical Definition of the "55" Personality 

The symptomatic, clinical aspects of the overtly self-effacing per- 
sonality are easily described. The outstanding symptom is depression. 
These people are overtly anxious and unhappy. They exhibit guilt 
and self-depreciation. Doubt, rumination, and obsessive uncertainty 
are emphasized. Associated with this is an immobilized passivity. 

They are not active or self-confident. They are not assertive or 
reasonable. They do not challenge or compete with the clinician. 

Their interpersonal impact on others involves weakness. They 
often admit their need for psychotherapy. They tend to make the 
clinician feel comfortable in his role because tHey readily assume the 
position of a patient. These are the patients who keep the clinics in 

These patients are often riddled by guilty, obsessive thinking. 
Hecht's investigation of the masochistic personality revealed that ob- 
sessive rumination (as measured by the MMPI) had an important diag- 
nostic relationship to self-effacement (2). This has been confirmed 
repeatedly by our own studies, which have revealed a correlation be- 
tween obsessive thinking and self-depreciating behavior. 

When a patient comes to the clinic emphasizing such messages as 
"People are mean to me," "I have done wrong," "I am unworthy," 
and "I am inferior," then the presenting operations of masochism can 
be suspected. The effect of this approach is to make the other one 
feel strong, slightly superior, perhaps, and initially supportive. The 
superior reaction of the clinician is often bound up in his therapeutic 
role so that he may not be aware that this response is being pulled 
from him. The untrained clinician is often provoked to sympathetic 
gestures. The more sophisticated diagnostician is usually struck by 
the force of the self-punitive superego. 

This brings us to another aspect of this personality type — the moral- 
istic quality of their self-reproaches. The masochistic, guilty patient is 
generally obsessed with matters of "right and wrong" and measures 
himself (to his own disadvantage) against his own strict ideals. This 
point is clearly demonstrated by the finding that the sector of the 
diagnostic circle which defines self-effacement is the farthest re- 
moved from the standard ego-ideal image of our culture. Their be- 
havior is rated in the HI sector of the circle — whereas the ego ideal is 


invariably located in the opposite sectors. These patients are dissatis- 
fied with themselves, and this is, of course, related to their relatively- 
high motivation for psychotherapy. 

In their social demeanor these patients typically tend to be silent, 
fearful, and unsociable. In the moderately self-effacing person this 
may be seen as a modest reserve. In severe cases it becomes a marked 

Obsessive Neurosis and Selj-Ejfacement 

The self-effacing personality manifests the symptoms of depression, 
immobilization, and ruminative self-doubt. Patients whose overt inter- 
personal behavior is masochistic or self-derogatory are often given the 
standard psychiatric diagnosis of obsessive neurotic. In Chapter 12 
common clinical diagnostic categories were compared to interpersonal 
types. Evidence was cited which showed that the HI sector of the 
circle at Level II was related to the familiar obsessive category. Re- 
search on Level I behavior has confirmed this finding. Patients who 
were diagnosed by fellow group patients as falling in the masochistic 
sector of the circle invariably manifested the symptoms of the obses- 
sional disorder. On the MMPI these patients have their highest scores 
on the depression and psychasthenia scales — which are generally seen 
as diagnostic of obsessional processes. 

Several correlation studies between MMPI scales and Level I be- 
havior have been reported (3). When the depression and psychas- 
thenia scales are correlated with the Level I-S vertical index, significant 
negative correlations with dominance are consistently obtained. De- 
pression and worry are related to passivity. 

There is considerable research evidence pointing to a relationship 
between the interpersonal security operation of masochism and ob- 
sessive symptoms and chnical diagnosis of obsessive neurosis. In addi- 
tion, there are some theoretical links between masochism and the ob- 
sessive process. When this relationship was first suggested by our 
data, there was considerable question on the part of the research staff, 
as well as the advising clinicians, as to the accuracy of tying masochism 
to obsessiveness. 

Subsequent diagnostic work has tended to confirm the relationship 
and has shed some light on its possible theoretical meaning. The link- 
ing factor seems to concern guilt and self-derogation. It is generally 
accepted that the obsessive symptoms — rumination, concern with 
right and wrong, self-doubt, etc. — are connected with guilt. So is 
masochism. It seems to make clinical and theoretical sense that self- 
effacement is the interpersonal expression, and obsessiveness the symp- 
tomatic expression of the same overt security operation. 


The Obsessive-Cofnpulsive Phenomenon 

The relationship between the interpersonal security operations of 
self-effacement and the standard symptomatic diagnosis of "obsessive" 
raises an interesting terminological issue. Obsessions have to do with 
persistent ideas, intellectual preoccupations, doubts, worries, guilty 
thoughts. These generally lead to inhibition of action — expressive, 
spontaneous action in particular. Obsessiveness is typically accom- 
panied by indecisiveness and depressive immobilization. The inter- 
personal correlate of obsessiveness is modest passivity and self-punitive 
timidity. Worried rumination communicates the interpersonal mes- 
sage, "I am unsure, fearful, self-doubting." 

Compulsions have generally been distinguished from obsessions. 
Compulsions are repetitive activities, e.g., promptness, orderliness, 
precise activity, disciplined behavior. Compulsions often have an 
interpersonal impact quite different from obsessiveness. Compulsive 
individuals are often not indecisively immobilized; they expend a great 
deal of energy in exact, demanding action. They often communicate 
not an interpersonal message of doubt or fear, but, on the contrary, 
one of righteous self-satisfaction, pedantry, and superiority. It seems 
in some cases that when compulsions are successfully executed they 
express the opposite interpersonal meaning of obsessive behavior. 

Obsessive and compulsive behavior are traditionally linked in 
psychiatric terminology. The terms are often used synonymously. 
In most diagnostic texts the two are considered together and a para- 
doxical mixture of symptomatic cues is lumped together. Worried 
self-doubt and pedantic superiority are often cited together as diag- 
nostic cues for the same personality type. From the standpoint of 
descriptive or symptomatic psychiatry these inconsistencies do not 
appear too striking; but when they are viewed from the position of 
interpersonal theory, the paradoxical and dichotomous nature of the 
obsessive-compulsive syndrome comes sharply into focus. 

The interpersonal meaning of successful compulsivity is, "I am 
right and superior." The interpersonal meaning of pure obsessiveness 
is, "I am wrong and unsure." 

The general practice of combining these two opposing security 
operations is a confusing and inefficient terminological practice. The 
functional meaning of rigid compulsivity is quite different from that of 
pure obsessiveness, and different from both of these are the many 
cases which show alternations of both behaviors. Obsessive-compul- 
sive is hyphenated because the two elements appear to be opposing, 
dichotomous factors; they are diametrically different ways of han- 
dling guilt and weakness. Sado-masochism is another familiar hyphe- 


nated term in psychiatry. These two elements are also Hnked because 
they are diametrically opposite ways of dealing with hostility. Clini- 
cians seem to recognize, however, that although sadism and masochism 
are reciprocally related, the two words are not synonymous. It is of 
crucial importance to know which side of a sado-masochistic conflict 
is overt and which is underlying. The functional problems involved 
in getting an overtly sadistic personality into therapy are quite dis- 
tinct from those involving the overt masochist. The latter is often 
initially better motivated. 

It is useful to make the same distinction in the case of the obsessive- 
compulsive phenomenon — that is, to determine specifically whether 
a patient is presenting overtly as an obsessive, guilt-ridden, depressed 
person, or whether compulsive defenses are successfully operating. 
In the latter case the patient is outwardly active, more self-confident, 
and manifests a righteous, active fa9ade. 

Many cases seen in the psychiatric clinic show mixtures of obses- 
sive-compulsive symptoms. In the interpersonal language they may 
be guilty and self-effacing at Level I-M, but this may be seen as a 
temporary breakdown of a compulsive personality. Often Level I-M 
may be depressed and masochistic while the Level II self-description 
emphasizes managerial, responsible themes. This indicates that the 
compulsive defenses are weakening; guilt and weakness in the form 
of symptoms are breaking through. 

Arthur Kobler of the Pinel Foundation Hospital has added an im- 
portant qualification to the point being made in this section. He be- 
lieves that the distinction between the interpersonal implication of 
obsessive versus compulsive behavior may hold for the popular, ad- 
justing aspect of compulsivity. He states, however, that severe com- 
pulsive rituals — "driven actions with magical quality" — are closer to 
obsessiveness. The interpersonal theory would be in strong agree- 
ment with this statement because it interprets these bizarre rituals as 
diagnostic of the schizoid message, "I am different, queer, alienated." 
Since schizoid behavior in the interpersonal system falls next to ob- 
sessiveness on the diagnostic continuum, Kobler's valuable clarification 
seems to fit the "circle" theory. 

To summarize: The distinction between compulsive and obsessive 
behavior is functionally valuable. Pure compulsivity (where there is 
no breakthrough of the warded-off, underlying guilt) indicates inter- 
personal power, pedantry, and self-righteousness. Pure obsessiveness 
is associated with overt interpersonal passivity and humiUty. It is 
possible to use the hyphenated term obsessive-compulsive to refer to 
multilevel patterns of conflict, but the meaning (symptomatic and 
interpersonal) of the separate terms should be kept distinct. 



Research Findings Characteristic of the 
Self-Effacing-Masochistic Personality 

The characteristics of the masochistic personahty which have just 
been discussed are based on research findings of the Kaiser Founda- 
tion project. These have been described in other publications. Some 
of these findings will now be summarized. 

1. Patients who exhibit masochistic operations at Levels I and II 
do not tend to have psychosomatic disorders, except for the overtly 
neurotic dermatological symptom groups (acne, seborrheic dermatitis, 
and psoriasis) . 

2. Psychosomatic patients do not present self-punitive behavior in 
their overt operations, except for the above-listed skin disorders. 

3. Patients who express masochism at Level I tend to have MMPI 
profiles emphasizing obsessive (Pt), depressive (D), and passive (Mf) 

4. Self-punitive behavior at Level II is also related to the same 
MMPI scales. 

5. These patients tend to stay in psychotherapy longer than hys- 
teric, managerial, narcissistic, or psychosomatic patients. They tend 
to stay in therapy about the same length of time as schizoid, phobic, 
and psychopathic personalities. They belong to the well-motivated 
group of patients. 


Percentage of Self-Effacing-Masochistic Personalities (Level I-M) 
Found in Several Cultural Samples 

% of Self-Effacing-Masochistic 

Institutional or Symptomatic Sample 



Psychiatric Clinic Admission 



College Undergraduates 


University Psychiatric Clinic 



Aliddle Class Obese Patients (Female) 


Neurodermatitis Cases 



Overtly Neurotic Dermatitis Patients 



Self-inflicted Dermatitis Patients 


Unanxious Dermatitis Patients 


Individual Psychotherapy Patients 



Hypertensive Patients 


Ulcer Patients 


Medical Control Patients 


University Counseling Center 


University Graduate Students (Male) 


Stockade Prisoners (Male) 



Hospitalized Psychotic Patients 



Officers in Military Service 





6. They tend to be consciously disidentified with their mothers. 

7. They tend to be consciously disidentified with their fathers. 

8. They tend to be consciously disidentified with their spouses. 

9. The masochistic personality is found most frequently in certain 
institutional and cultural settings (see Table 18). Masochists appear 
more often in psychiatric samples and rarely occur in psychosomatic 
or normal samples. One exception to this statement — certain neuro- 
dermatitis groups are more often masochistic (at Level I) than any 
other psychosomatic sample. 

10. The percentage of self-effacing personalities (defined by Level 
II-C) found in various samples is presented in Table 19. 


Percentage of Self-Effacing-Masochistic Personalities (Level II-C) 
Found in Several Cultural Samples 

% of Self-Effacing-Masochistic 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 
Hospitalized Psychotic Patients (Male) 
Group Psychotherapy Patients 
Individual Psychotherapy Patients 
Overtly Neurotic Dermatitis Patients 
Self-inflicted Dermatitis Patients 
Unanxious Dermatitis Patients 
Medical Control Patierits 
Ulcer Patients 
Hypertensive Patients 
Middle Class Obese Patients (Female) 

Neurotics and neurodermatitis patients express the most masochism. 
Psychosomatic and normal groups the least. It is of interest that the 
psychotic group manifests considerably less self-effacement than the 
neurotic samples. The impHcations of these findings are discussed in 
Chapters 23 and 24. 


1. Grotjohn, M. Jake Gimbel lectures, University of CaUfomia, 1955. 

2. Hecht, Shirley. An investigation into the psychology of masochism. Unpub- 
lished doctor's dissertation, University of California, 1950. 

3. Leary, T., and H. Coffey. The prediction of interpersonal behavior in group 
psychotherapy. Psychodrama gr. psychother. Monogr., 1955, No. 28. 

4. Menaker, Esther. Masochism— a defense reaction of the ego. Psychoanal. Quart., 
1953, 22, No. 2, 205-20. 
























Adjustment Through Docility: 
The Dependent Personality 

This chapter is concerned with those individuals who present in their 
approach to the clinic a fa9ade of dependent, docile conformity. This 
is the ''66'' personality type. The interpersonal message it conveys to 
others is, "I am a meek, admiring person in need of your help and 

The moderate form of this security operation is expressed as a re- 
spectful or poignant or trustful conformity. In its maladaptive inten- 
sity it is manifested as a helpless dependency. These subjects in their 
interpersonal reflexes avoid the expression of hostility, independence, 
and power. 

The Purpose of Docile Conformity 

Human beings utilize these security operations because they have 
found that they are least anxious when they are outwardly relying on 
or looking up to others. Some individuals employ these reflexes in 
their relationships with everyone they contact. Others assume this 
role when they assume it to be called for by the situation. They act 
helpless and fearful when dealing with strong individuals, authority 
figures, and the like. 

Many patients automatically assume this role in approaching medi- 
cal or therapeutic agents. The doctor-patient relationship is loaded 
with dependency implications. Most patients manifest a certain 
amount of helpless trust in coming for diagnosis. The normative sta- 
tistics employed in the interpersonal diagnostic grids are based on 
large samples of clinic patients. In this chapter therefore we shall be 
describing those individuals who express more dependency than the 
average clinic visitor. We have isolated these persons who seem to go 
out of their way to pull sympathy, help, and direction from others; 



who use their symptoms to communicate a helpless, painful, uncertain, 
frightened, hopeful, dependent passivity. 

Now many of these patients exert this interpersonal pressure in the 
clinic but may act quite differently in other situations. They may be 
fairly independent in certain social interactions. At other levels of 
behavior they may be less phobic and docile. The fact that they pre- 
sent dependence as their calling card to the clinic is an indication that 
this is the functionally critical point at which to begin the diagnostic 
evaluation. This defines their initial motivation, their first line of de- 

Whenever we observe this security operation, we tend to assume 
that the patient has come to employ docile dependence as a means of 
handling anxiety in this type of situation. Further investigations may 
reveal the flexibility or rigidity of this behavior and may indicate 
that opposing motives exist at other levels or in other situations. 

The Effect of Docile Conformity Upon Others 

Docility pulls strong, helpful leadership from others. Dependence 
provokes nurturance. In the language of the circle, ''JK pulls AP and 
NO from others." 

If a person acts in a poignant, helpless, respectful manner, he trains 
others to offer help, advice, and direction. He who asks tends to get 
taught. These subjects tell others by means of their reflexes that they 
are weak-and-friendly. They thereby provoke others to be strong- 

These reciprocal tendencies do not occur inevitably but within 
probability limits. Some punitive individuals react with stern disap- 
proval to dependence in another. Severe masochists are unable to ex- 
press nurturance even though the other is exerting intense dependent 
pressure. In general, however, docile individuals tend to be most com- 
fortable when they are involved with strong, responsible individuals. 
Nurturant people naturally seek admiring, trustful individuals who 
will respond to and need their help. The docile phobic person tends to 
irritate the rebel and to threaten the counterphobic; he does not gen- 
erally integrate durable relations with these individuals. Close sym- 
biotic ties link the meek, admiring (JK) individual to respected, help- 
ful (APNO) partners. 

These reciprocal situations hold for brief encounters as well as 
durable interpersonal pairings. Poignant, tearful helplessness in the 
first few seconds of an interaction provokes tenderness and guidance 
from another. Patients who present these reflexes in an initial psychia- 
tric interview generate forces which may tend to pull assurance from 
the clinician. Whenever the clinical interviewer finds himself un- 


usually inspired to help, to promise, to reassure, to explain, to do some- 
thing to relieve anxiety and tears, he will generally find that he is deal- 
ing with security operations of dependent docility. 

The effect of JK behavior is, therefore, to train the "other one" to 
assume a strong, friendly role. Circular chains of interaction, of course, 
develop. The respected, responsible, nurturant person in turn presses 
the dependent person to increased dependence. Where these sym- 
biotic tendencies are uncomplicated by underlying conflicts on the 
part of either partner, a most comfortable durable relationship de- 
velops. The passive son attached to a strong nurturant mother pro- 
vides a typical example of this process. The docile, adoring wife 
dutifully tied to a responsible, managerial husband is another. 

Where the docility is intense and all other reflexes are crippled or 
where underlying motives conflict with the overt dependence, then 
anxiety fails to be warded ofl^. This anxiety can be dealt with by in- 
creased helplessness, eventually leading to a fairly typical set of psy- 
chological symptoms. The neurotic expressions of severe dependent 
conformity will now be considered. 

Clinical Definition of the ''66" Personality 

The defensive operations of docile conformity, when employed in 
the intense maladaptive degree, result in a set of specific symptoms 
which are related to and a logical outcome of the tactics. 

The first clinical indication is helplessness and overt anxiety. Clini- 
cally this is generally expressed as a marked depression. Fears, wor- 
ries, elaborate concern over physical or emotional discomfort are 

A most definitive sign of this personality type is the presence of 
phobias. The patient is fearful of events or experiences without any 
direct rational cause. Descriptive psychiatry of the last century has 
listed dozens of impressive-sounding hyphenated terms denoting the 
different phobic reactions. While it seems fruitless to recapitulate this 
list of descriptive labels, it seems worth while to point out that they 
generally refer to an irrational and inexplicable intense fear of some 
stimulus — fear of heights, fear of crowds, fear of being alone, etc. 

The theory of the interpersonal circle offers one possible rational 
correlation of these fears with docile-dependent operations. Fears, of 
course, tend to give the impression of weakness and helplessness. This 
pulls for help and support. 

But the fears of the phobic are, in essence, displaced fears. It is well 
known that the relatively innocuous stimuli avoided by these patients 
generally stand for more directly intimate interpersonal figures who 
are covertly feared. We recall that Little Hans's panic about horses 


was related to certain unconscious perceptions of his father and 
mother. (2) A repressive tendency is at work here. The patient can- 
not directly attribute hostile, dangerous motives to real, known figures 
but unconsciously displaces these motives to figures or stimuli which 
are vague, and psychologically distant. One effect of this is to allow 
the patient to preserve a consciously conforming, docile relationship 
with close figures against whom he may feel negative, rebellious emo- 

At this point the theory of the interpersonal circle can be intro- 
duced. The points / and K which define the phobic personality are 
midway between weakness (HI) and conventional agreeability (LM). 
On the circular diagnostic continuum, the phobic is related on the one 
hand to the obsessive and on the other to the hysteric. Clinically this 
suggests that phobics combine fears and obsessions on the one hand 
with a conventional, repressive, bland tendency to see family mem- 
bers and intimates as sweet and loving. The MMPI pattern for the 
phobic personality involves peak scores on depression, psychostenia, 
and hysteria. This tends to confirm the clinical impression of a per- 
son who is unhappy, anxious (D), worried, and fearful (Pt), and at 
the same time blandly repressive (Hy). He is afraid, but he does not 
know what he fears. He is helpless and weak within the context of 
docile, naive conformity. 

So far we have emphasized the phobic symptoms of the docile-de- 
pendent personality. The point has been made that inexplicable fears 
seem to fit nicely the mixture of weakness plus bland conventionality 
which characterizes this personality type. There are other symptoms 
which allow the patient to be helpless, depressed, and anxious, and to 
maintain a conforming conscious picture of self and others. Diffuse 
physical symptoms, for example, have the same psychological implica- 
tions. They tend to be typical of, and partially diagnostic of, the 
docile phobic personahty. 

We are considering here symptoms which seem to be physical ex- 
pressions of anxiety and tension; insomnia, transient digestive com- 
plaints (e.g., "butterflies in stomach" and nausea or bowel reactions in 
response to stress) and transient circulatory symptoms (blushing, faint- 
ing behavior, cardiac responses to stress, etc.). Many diffuse hypo- 
chondriacal concerns have docile-dependent overtones. 

The interpersonal function of these symptoms is to present a pic- 
ture of a worried, distressed person in need of help because of symp- 
toms which have an indirect emotional significance. The patient suf- 
fering from diffuse physical symptoms often does not complain spe- 
cifically or directly about his interpersonal problems or those of his 
intimates but displaces much of his concern onto areas which are 


psychologically more distant and much more indirect. Again, he is 
anxious but he does not know why. 

The third and most pathonomonic set of symptoms characteristic of 
the docile-dependent personality includes the manifestations of overt, 
free-floating anxiety. When a patient comes to the clinic openly ex- 
pressing signs of weakness, discomfort, concern over self (e.g., tears, 
fidgeting, fearful behavior), then the security operations of phobic 
conformity may be suspected. The interpersonal message expressed 
by these tactics seems to be: "I am a distressed, weak, unhappy person 
in need of your help and direction." This and the preceding generali- 
zations refer to the generalized or pure or consistent case. Alany pa- 
tients manifest alternations of behavior in a diagnostic interview. They 
may initially exhibit interpersonal reflexes of self-confident superior 
strength and then lapse suddenly into fearful, tearful behavior. In 
this case the hypothesis of intense phobic-counterphobic conflict 
would perhaps be considered. 

Relationship of Docile Conformity to Standard 
Psychiatric Diagnosis 

The preceding section has suggested that the interpersonal traits of 
overt docile dependence are related to certain clinical symptoms. 
These were anxiety, phobias, and diffuse physical symptoms. 

Patients who employ these operations and manifest these symptoms 
can be given five difl^erent standard psychiatric diagnoses. The diag- 
nostic label used is generally determined by the kind of symptoms 
which characterize the patient. 

1. The term anxiety neurosis generally defines a docile-dependent 
personality. Malamud describes this diagnostic type as follows: 

Clinically, this disturbance expresses itself in attacks of vague, unexplained 
but intense fear which, at least in the beginning of the disease, does not seem 
to be attached to any particular object. It can best be described as being near 
to a normal fear of a vital danger, but is different from it in that no such dan- 
ger is present and, in most cases, not even imagined to be present. The con- 
comitant symptoms are usually of the same kind as found in real fear— a kind of 
paralyzed state of the musculature, cold shivers, a sense of pressure in the head 
and precordial regions, profuse cold sweating, palpitation of the heart, and at 
times relaxation of the sphincters. As time goes on the attacks may be con- 
sciously associated with some of the concomitant symptoms. The person may 
develop the fear that his heart may stop, that something will burst in his head, 
or that some serious disease is developing in his gastro-intestinal system. The 
concomitant symptoms may also assume the controlling feature of the picture, 
and thus instead of pure anxiety attacks we may have tachycardia, alternating 
constipation or diarrhea, dizziness, or even vertigo, and others. (3, p. 853) 

The patient who complains of these symptoms usually approaches 
the clinician in a dependent manner, seeking relief and help. 


2. The term phobic is also used to describe docile patients. Again 
the interpersonal implication of the fearful state is that the patient is 
a weak, helpless person. There is considerable overlap in the descrip- 
tions of anxiety neurotics and phobics. Notice in Malamud's definition 
the emphasis placed on fears. The difference between anxiety neurosis 
and phobia seems to involve superficial descriptive aspects of the con- 
tent of what is feared. Both types seem to describe the same generic 
personality syndrome. 

3. In the Freudian literature a similar overlapping of terms occurs. 
In defining anxiety hysteria Fenichel states that "the anxiety is spe- 
cifically connected with a special situation, which represents the neu- 
rotic conflict. (1, p. 194) In discussing "the choice of the specific 
content" of the fears in anxiety hysteria, Fenichel (1, p. 195) moves 
immediately to a consideration of phobias, and it is clear that he con- 
siders phobias the characteristic symptom of the anxiety hysteric. 

It appears that the terms anxiety and phobic as used in diagnostic 
labels are descriptive and symptomatic. The value of these terms for 
nosology is limited. One is led to question (1) the usefulness of 
descriptive diagnostic labels and (2) the proliferation of these over- 
lapping terms. 

4. There is a fourth standard diagnostic category which is related 
to the docile-dependent personality. This is the term neurasthenic. 
Malamud gives a description of this condition: 

In its pure form it is characterized by feelings of physical and mental in- 
adequacy, complaints of fatigability without adequate exertion, paresthesias in 
the back of the neck, and a sense of general weakness. In the more chronic and 
severe forms of this disturbance the patients usually describe themselves as 
mental and physical "wrecks." They cannot concentrate on any activity, they 
wake up in the morning feeling exhausted, "fagged out," unable to get started 
on any work. Irritability, feeling of lack of sexual vigor at times amounting to 
impotence, and a vague sense of anxiety may complicate the picture. In contra- 
distinction to the anxiety neuroses, these states are usually monotonously 
chronic without any great degree of variation and as is too frequently the case 
the patient seeks for help only after long duration of the symptoms. (3, p. 854) 

The similarity of neurasthenia to anxiety neurosis is apparent in 
this description and is, in fact, recognized by Malamud. It seems that 
the differentiating factor is chronicity — a dubious reason for retaining 
a nosological category. 

5. A fifth diagnostic term which usually defines the overtly de- 
pendent personality is hypochondriasis. Diffuse physical symptoms 
and worry about bodily functions can serve as a defense in many types 
of maladjustment. Often these concerns operate in very sick patients 
to ward off psychotic processes. It is safe to say that in any hypo- 
chondriacal condition, whatever the underlying problem, displace- 


ment and repressive processes are at work. At the level of presenta- 
tion to the clinic the interpersonal implications of hypochondriasis are 
(1) dependence and need for help and (2) some tendency to displace 
negative emotions onto physical reactions. These two factors are 
characteristic of the personality type we are discussing in this chapter. 

Research Findings Characteristic of the ''66'' Personality 

This section presents a summary of some of the empirical studies 
accomplished on the phobic personality at the Kaiser Foundation re- 

1 . Patients who present docile conformity in their overt operations 
do not tend to have the psychosomatic symptoms of ulcer, hyperten- 
sion, or neurodermatitis. (Although they do not manifest organ 
neuroses, they do tend to complain of diffuse physical symptoms of 

2. Docile dependency at Levels I and II is related to depression 
(D), ruminative worries (Pt), and naive blandness (Hy) on the 

3. These patients tend to be initially well motivated for treatment, 
remaining in treatment for an average of eleven interviews. Pure 
phobics (i.e., without underlying ambivalence) remain in treatment 
an average of twenty times, that is, longer than any other diagnostic 
group. They are solidly docile and dependent. Conflicted phobics 
(i.e., with underlying hostility, strength, or conventionality) on the 
contrary do not remain in treatment, being seen on the average of 2.6 
sessions. This dramatic reversal of the pure and the more ambivalent 
cases points up the necessity of fitting the variability dimension into 
the diagnostic picture. 

4. Docile subjects are on the average ambivalent in their conscious 
identification with parents. They are not so disidentified as the 
psychopaths, schizoids, and obsessives. They are less close to their 
parents than the conventional and responsible personality types. 

5. They are similarly about in the middle on the variable of marital 
identification, being closer to their marital partners than the uncon- 
ventional diagnostic groups and less close than the conventional. 

6. Docile patients consciously describe their parents as being con- 
ventional, agreeable, and somewhat nurturant people. This reflects a 
conforming attitude to parents and places them close to the hysterical 
and psychosomatic patients. Phobics picture themselves as weaker 
than the latter two personality types, but share their conventional per- 
ception of parents. 

7. Phobics emphasize nurturance in their conscious description of 
marital partners more than any other diagnostic group. They are 


themselves dependent and marry people whom they see as strong and 

8. They are therefore not identified with their spouses but report 
reciprocal "needs help-gives help" marital relationships. 

9. On the Naboisek study of interpersonal misperception, phobics 
(when combined with obsessives) seem to be the most accurate of any 
diagnostic type. They correctly perceive the strong to be strong and 
the weak to be weak. They manifest misperception only in the case 
of the hostile persons, to whom they erroneously attribute more weak- 
ness than hostility. Docile dependent patients seem to be thrown off 
by aggressiveness in others, preferring to see this as weakness. This 
may reflect an avoidance of the same interpersonal themes they avoid 
in their overt behavior. 

10. The phobic personality is found most frequently in certain in- 
stitutional and cultural settings. Docile people (Level I-M) do come 
to the psychiatric clinic for help. They are not found as frequently 
in normal nonclinical settings (see Table 20). 


Percentage of Docile-Dependent Personalities (Level I-M) 
Found in Several Cultural Samples 

% of Docile-Dependent 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 537 12 

College Undergraduates 415 2 

University Psychiatric Clinic 133 10 

Middle Class Obese Patients (Female) 121 5 

Overtly Neurotic Dermatitis Patients 31 10 

Self-inflicted Dermatitis Patients 57 5 

Unanxious Dermatitis Patients 71 3 

Group Psychotherapy Patients 109 11 

Individual Psychotherapy Patients 49 16 

f lypertensive Patients 49 5 

Ulcer Patients 43 5 

Medical Control Patients 57 5 

University Counseling Center (Male) 93 2 

University Graduate Students (Male) 39 5 

Stockade Prisoners (Male) 52 6 

Hospitalized Psychotic Patients 28 21 

Officers in Military Service 39 

Total 1903 

11. Docile-dependent patients tend to be assigned to individual 
psychotherapy. They are second only to the obsessives in the per 
cent referred to and remaining in individual treatment. They are not 
referred as frequently to group therapy. They do not remain in group 


psychotherapy as frequently. Four other diagnostic types supply 
more patients who remain in groups. The phobics top only the hys- 
terics, psychopaths, and narcissists in percentage of group therapy 
numbers. This is probably due to the following facts: The phobic 
tends to be quite ambivalent about treatment in general. He wants 
help but not necessarily psychological exploration. When he is seen 
individually, this ambivalence about motivation and commitment to 
therapy can be made the focus of attention and dealt with directly. 
The docile patient is more likely to feel comfortable in a two-way 
doctor-patient relationship. The dependent operations work more 
smoothly. In a group the ambivalent motivation is very easy to over- 
look. The patient does not have a single comfortable situation of a 
nurturant therapist but is thrown into interaction with several other 
patients and personality types. Underlying ambivalences can be 
intensified and the phobic often drops out of the group. 

12. The frequency of docile-dependent subjects at Level II-C is 
presented in Table 21. It will be noted that this personality type is 
again most numerous among hospitalized psychotics. The second most 
frequent occurrence of this personality type is in the individual therapy 


Percentage of Docile-Dependent Personalities (Level II-C) 

Found in Several Cultural Samples 


of Docile-Dependent 

Institutional or Symptomatic Sample 



Psychiatric Clinic Admissions 



Hospitalized Psychotic Patients (Male) 



Group Psychotherapy Patients 



Individual Psychotherapy Patients 



Overtly Neurotic Dermatitis Patients 



Self-inflicted Dermatitis Patients 



Unanxious Dermatitis Patients 



Medical Control Patients 



Ulcer Patients 



Hypertensive Patients 



Middle Class Obese Patients (Female) 



Total 781 

Therapeutic Handling of the Phobic Personality 

Some phobics express underlying counterphobic power; others 
show sweet hysterical conventionality in their "preconscious" expres- 
sions. Entirely different therapeutic results may be expected from 
these varied multilevel patterns even though they are all presented 
overtly in the same way. Generalizations about the phobic personality 


must therefore be considerably limited, although it is possible to ex- 
amine some of the implications of the docile fagade which seems to 
characterize most phobic patients. 

The first characteristic worth noting is their apparent readiness and 
eagerness for psychiatric help. The word help is used here in contrast 
to the word treatment — for many phobics are not at all eager for ex- 
tended therapy. Because of their dependence, their admission of fear 
and weakness, they give the appearance of being highly motivated and 
cooperative patients. This appearance is often misleading. 

Phobic patients in their underlying levels have their share of all the 
sixteen generic interpersonal motivations. They have as much (or 
perhaps more) ambivalence and conflict as any other overt personality 
type. The deceptive factor here is the fact that they have a fa9ade of 
cooperative, passive docility. This often lulls the imperceptive clini- 
cian into the expectation that the patient is wholeheartedly involved 
in the treatment plan. If questioned, the phobic may appear to be in 
complete conformity with the program outhned by the clinician. 
Here we think of the typical and familiar phrase, "I'll do anything 
you suggest, Doctor." 

Thus the ambivalence and conflicting motivation which we expect 
in almost every patient tends to be easily overlooked in the case of the 
docile phobic. 

In dealing with most other overt personality types, the intake diag- 
nostician is automatically led to look for ambivalence or conflict. In 
the case of the schizoid patient some partial abandonment of his dis- 
trustful operations is required in order for him to express the collabora- 
tive feelings involved in a commitment to therapy. Similarly, a strong, 
self-confident counterphobic patient must make some admission of 
weakness and need-for-help if he is to commit himself to treatment. 
In the case of the docile phobic patient, ambivalence is often present 
but can be easily overlooked because of the superficial eagerness of 
these patients to please and conform to the clinician's suggestions. For 
this reason these patients tend to present tricky and confusing prog- 
nostic problems. In the early days of the clinical training program at 
the Kaiser Foundation clinic, phobic patients were often assigned for 
therapy to novitiate interns. The reasoning was: "These patients are 
anxious, cooperative, well-motivated, and not too distrustful, and are 
thus excellent patients for the beginning therapist." 

This generalization has proved optimistic. We have found it to be 
difficult to predict the clinical course of an overtly docile patient. 

In a preceding section it has been pointed out that phobics on the 
diagnostic continuum fall between hysterics and obsessives. Func- 
tionally, this means that they tend to combine punitive self-deprecia- 



tion and bland naivete. The repressive hysterical element often leads 
phobics to an early departure from the clinic as soon as they sense that 
therapy is not a magical cure but rather a process of realistic self- 
evaluation. The latter, of course, is quite alien to the repressive opera- 

We have found that many phobics can present an initial facade 
which involves severe anxiety, marked conformity to treatment plans, 
and apparent motivation for therapy. The underlying motivation may 
involve other interpersonal operations and might predict an early 
"repressing" out of therapy, or the development of severe feelings of 
distrust and isolation, etc. Phobics may often resist (in a conciliatory 
manner) the clinician's attempt to clarify their motivation — particu- 
larly if they sense that a reproach or criticism is impUed. This is gen- 
erally followed by a reaffirmation of their willingness to conform to 
the "doctor's orders." A supportive and sympathetic explanation of 
the phobic's motivation will often allow the prtient to express his un- 
derlying doubts, or fears, or critical resistance to psychotherapy. 

Again it must be recalled that these comments are limited by the 
multilevel variations which differentiate the 2,048 types who present 
overtly as docile-dependent. The temporal sequence of interpersonal 
behavior to be expected varies according to the configuration of the 
total personality. The therapeutic handling of overt phobics, there- 
fore, varies according to these differences. 


1. Fenichel, O. The psychoanalytic theory of neurosis. New York: Norton, 1945. 

2. Freud, S. Analysis of a phobia in a five-year-old boy. Collected papers. Vol. 3. 
London: Hogarth Press, 1948. 

3. Malamud, W. The psychoneuroses. In J. McV. Hunt (ed.), Personality and the 
behavior disorders. New York: The Ronald Press Co., 1944. 


Adjustment Through Cooperation. 
The Overconventional Personality 

Conventional, friendly affiliation with others is the mode of adjust- 
ment discussed in this chapter. This is the "77" personality type. We 
shall be discussing those individuals whose overt security operations 
involve agreeability, and who strive to be liked and accepted by others. 

Adaptive Forms of the Conventional Personality 

Extroverted friendliness is the adaptive form of "this generic secur- 
ity operation. The individuals who utilize these interpersonal reflexes 
seem to be comfortable when they are evoking "good feelings" and 
establishing harmonious, amicable relations with others. 

They tend to seek satisfaction in sociabihty with others. Accepted 
values are important to them. They are more likely to cooperate, to 
go along with the conventional pattern, to compromise. External 
harmony is more important than internal values. They are less likely 
to emphasize a unique, original, or highly controversial point of view. 

Individuals who employ this interpersonal machinery with flexibil- 
ity are productive and valuable members of society. They are popu- 
lar, well-liked, and agreeable members of any group. They deal with 
social anxiety by friendly, amicable responses. 

This mode of adjustment is probably the highest stated ideal of our 
Western civilization. The loving, peaceable, brotherly person is given 
the most honored role in the ethical hierarchy. This is, it must be 
noted, a cultural ideal. The personal ideal of most individuals (as 
measured by the interpersonal system) clearly emphasizes a combi- 
nation of conventionality and strength. The pure loving person is the 
third most idealized figure, power and sympathetic responsibility 
being the ego-ideal values preferred by the individuals studied. The 
person who acts or describes himself in terms of cooperativeness and 



friendliness seems to be attempting to meet the stated rather than the 
real cultural conventional standard. 

Maladaptive Forms of the Conventiojial Personality 

Individuals who rigidly and inappropriately express agreeable, 
afEliative behavior are diagnosed as overconventional personalities. 
These are the persons who cannot tolerate any critical or strong or 
guilty behavior in themselves. They continually strive to please, to 
be accepted, to establish positive relations with others. 

It is difficult to describe these security operations because the Eng- 
lish language has a scarcity of words denoting this condition. It was 
pointed out in Chapter 2 that our English dictionaries do not contain 
terms defining the state of being overaffectionate or too friendly or 
overcooperative. The notion that a person can be maladaptively sweet 
is apparently alien to our culture. 

Thus we face the dilemma of describing persons for whom there 
exists no ready-made, common terminology. We have had to meet 
this problem (in our empirical studies) by hyphenated words or ex- 
tended phrases denoting the person who is abnormally and rigidly 

The items on the interpersonal check list which designate this con- 
dition are: 

L M 

too easily influenced by friends fond of everyone 

will confide in anyone likes everyone 

wants everyone's love too friendly 

agrees with everyone loves everyone 

The interpersonal behaviors which diagnose these security opera- 
tions involve the compulsive, repetitious expression of affiliative be- 
havior. These individuals smile, agree, collaborate, conciliate. They 
are extroverted and outgoing to an intense degree. They are so com- 
mitted to conventional responses that they forfeit originality and in- 

External values and approval from others dominate their social in- 
tercourse. Bland, often naive, uninsightful behavior is the inevitable 
correlate. Gross misperceptions of social reality characterize their 
approach. They just cannot see hostility or power in themselves. 
They avoid feelings of depression. A rigid overoptimism is quire 
typical. They often misperceive the interpersonal behavior of others 
and tend to saturate all their social exchanges with affihative motifs. 

The maladaptive aspects of these security operations are obvious. 
Like any set of crippled reflexes, the repetitious and inappropriate ex- 


pression of positive feelings lends to a general restriction of personality. 
Their repertoire of responses is narrow. They are limited only to the 
conventional actions and perceptions. They forget or misinterpret 
other kinds of behavior. They seem to fear being individuals. Their 
imagination and creativity is lost in the attempt to be acceptable or 
to be liked. A sterile conventionality or a self-satisfied piousness re- 

The Purpose of Overconventional Behavior 

The security operations of conventional agreeability are employed 
to ward off anxiety. These individuals are uncomfortable in the 
presence of hostile, unhappy, or power-oriented feelings. They avoid 
these responses. 

The overconventional person apparently has learned that he can 
reduce anxiety and gain heightened self-esteem by means of opti- 
mistic blandness. He has discovered that acceptance and approval 
from others can be won by means of friendly operations. He feels 
safe, comfortable, secure when he is employing these protections. 

When the reality situation involves unconventional behavior or 
threatens their optimistic (and often shallow) approach, these sub- 
jects become upset. They may strive to handle the situation by in- 
creased optimism and sociability. If these maneuvers fail, they tend to 
get out of the anxiety-provoking field. Psychiatric evaluation and 
self-exploration are, of course, among the most threatening events 
faced by this personality type. 

Their unique methods for handling the anxiety aroused by psychia- 
tric referral will be discussed below. 

The Effect of Overconventional Behavior 

Friendly agreeability tends to provoke approval and friendliness 
from others. In the systematic language, LM pulls MN from others. 
The extroverted, optimistic person trains others to like him. Co- 
operativeness induces a reciprocal positive response in others. 

These reciprocal relations are, of course, part of the folklore of our 
culture. The Dale Carnegie texts and the salesmen's manuals have 
pointed out the effect of the "positive approach," and our empirical 
studies have tended to confirm these bromides. 

A qualifying remark must accompany these generalizations. The 
principle of reciprocal interpersonal relations is a probability state- 
ment. It tends to hold most of the time. There are many cases where 
it does not work. LM does not always pull MN. 

The maladaptive intensity of the response provides a special case. 
Many situations call for anger or sorrow or power. If the overcon- 


ventional person is unable to respond appropriately, his attempts to 
win approval may fail. 

The personality of the "other one" is another important factor. 
If the alter in any social interaction tends to respond with a different 
inteq^ersonal reflex, then the ability of the overconventional person 
to pull approval is limited. Skeptical individuals can be infuriated by 
overoptimism in another. Power-oriented individuals may see co- 
operative agreeability as a form of docility and an invitation for them 
to increase their bossy reflexes. 

The selectivity of interpersonal relationships enters the picture at 
this point. Overconventional people tend to avoid persons and places 
which threaten their fa9ades. The "sicker" or more restricted the 
person is, the less able he is to tolerate differences which raise anxiety. 

Thus it often transpires that bland overfriendly persons tend to 
gravitate towards other agreeable, optimistic, pious, conventional peo- 
ple and do not tend to seek out antisocial or highly original partners. 

In cases where negative feeUngs are involved, these patients charac- 
teristically resolve the situation by the maneuver of "going along" 
with the feelings of their in-group. If the group to which they con- 
form is angry, they can be angry; but the hostihty is directed against 
an out-group figure and it is usually not expressed directly. These sub- 
jects can be very critical of an out-group person who is not present. 

The generalizations made in this section require qualification. 
Multilevel variations and conflicts provide new complications. Some 
patients with overconventional fagades tend to have underlying feel- 
ings which involve less-conventional themes (such as masochism or 
sadism). These "preconscious" tendencies may lead them to become 
involved with individuals who are unloving and unconventional. 

Even with the qualifications introduced by multilevel conflicts 
and by the personality of the "other one" the general principle of 
reciprocity holds as a low-order probability statement. Patients with 
low scores on nonconformity (i.e., the F scale on the MMPI) see them- 
selves as loving and cooperative and are seen in the same positive 
fashion by fellow group therapy members. 

Clinical Manifestations of the "77" Personality 

There are several cHnical characteristics of the overconventional 

First it should be noted that this personality type is not a common 
visitor to the psychiatric clinic. The essence of the psychiatric process 
is self-examination and an analysis of one's own unique patterns of 
living. The essence of the bland, friendly overagreeable mode of 


adjustment is the inhibition of one's uniqueness and one's individual 
feelings and the emphasis on external values. 

The overconventional person does not come to the psychiatric 
clinic because of a dissatisfaction with self or a desire for self-exami- 
nation. He is not depressed. He does not complain of internalized 
emotional problems (e.g., guilt, distrust). 

The specific symptomatic picture can vary, but the interpersonal 
message of bland self-acceptance is usually present. 

There are three reasons which bring these patients to a psychiatric 
clinic. These are (1) generalized "nervousness" or anxiety, vaguely 
defined and not tied to emotional causes; (2) physical symptoms, often 
with a direct symboUc meaning; (3) complaints about the behavior of 

The first of these complaints — generalized nonspecific anxiety — is 
the most common symptom, so typical that it is quite diagnostic. 
These patients use the words tension, nervousness, and cnixiety in their 
self-descriptions. The significance of these particular terms is that 
they have a relatively vague quality. They designate a symptom 
which is not tied to a specific, recognized emotional problem. The 
patient does not know why he is anxious. He is not depressed or fear- 
ful. This differentiates the overconventional from the phobic personal- 
ity. The latter is unhappy and sees himself as weak and timid. The 
overconventional person comes to the clinic because of anxiety which 
is described as a phenomenon quite removed from his personality. 

This symptom of vague tension can generally be traced to an inter- 
personal trauma or friction in the patient's life. Pressure is being put 
on the patient to react in a negative way (hostile or defeated). The 
overconventional person cannot handle these situations appropriately. 
He strives not to recognize the emotions which they arouse in him. 
His rigid attempts to misperceive and deny negative feelings in him- 
self and others seal off the emotional meaning and leave him only 
with intense anxiety. The threat of his own negative feelings (usually 
provoked by the traumatic external pressure) is the most intolerable 
experience for this personality type. He comes to the clinic, needless 
to say, not consciously desiring to have the cover removed from his 
misperceptions and negative emotions but to have the anxiety removed. 
When these patients sense that psychotherapy might threaten their 
bland denial they clearly express their disinterest in treatment. 

These patients present particularly pathetic pictures when they 
arrive at the psychiatric clinic. Their fear of their own negative feel- 
ing brings on the tension, but psychiatric interviews tend to arouse 
exactly the same anxiety. Caught between the pain of the illness and 


the pain of the cure, they usually handle this dilemma by intensifica- 
tion of their favored security operations; that is to say, they attempt 
to re-establish their bland, optimistic protections and move themselves 
out of the therapeutic situation. 

The complaint of diffuse tension can thus be seen as a symptom 
external to the patient's view of his own character structure. They 
come to the clinic seeking relief from this isolated symptom and not 
psychological explanation. This extended discussion of one sympto- 
matic presentation has been outlined in detail for two reasons. First, 
it is important for the clinician to recognize the fact that the com- 
plaint of tension or nervousness is not attached to the patient's con- 
ception of himself as a person. The intensity of the anxiety may make 
these patients appear to be well motivated for therapy. The bland, 
conventional nature of their security can be revealed by sensitive inter- 
viewing and is picked up very clearly in the Level I and Level II tests. 
Failure to distinguish this difference may lead to a breakdown in 
communication and the patient's flight from the clinic. The second 
important aspect of this syndrome is its frequency. Over 50 per cent 
of the overconventional patients seen in the Kaiser Foundation men- 
tion the vague, nonspecific terms tension, nervousness, or anxiety in 
describing their reasons for coming to the clinic. 

The second most frequent symptom mentioned by "77" patients 
involves physical complaints. These are often symbolic of unrecog- 
nized emotional conflicts. Headaches (which our clinicians believe 
to reflect underlying hostility) and menstrual complaints (believed to 
reflect sexual constriction) are probably the most common physical 
symptoms. Examination of the case material of overconventional pa- 
tients reveals that the great majority of the female patients are sexu- 
ally frigid. This is sometimes recognized but is rarely developed as a 
complaint, these patients being unalarmed about this condition. The 
physical symptoms classically characteristic of hysterical blandness 
(e.g., paralyses, amnesias, anaesthesias) are rarely seen in the Kaiser 
Foundation clinic. When they do appear they are not generally re- 
ported by the pure overconventional personality (77) but by severely 
or chronically disturbed patients with conflicted fagades. We think 
here of the schizoid-hysteric (47) or the masochistic-overconven- 
tional conflict (57). 

The third clinical characteristic of the "77" personality involves 
complaints about the behavior of other people. A sudden flare-up 
of marital trouble (previously unrecognized) is a common precipitat- 
ing event. The spouse may demand a divorce, thus breaking through 
the optimistic fa9ade and confronting the "surprised" patient with 
unpleasant emotions. Antisocial behavior on the part of a family mem- 


ber (delinquency, crime, sexual eccentricity) may bring about the 
same result. 

These situations confront the patient with emotions which his 
security operations have previously denied. It must be noted, how- 
ever, that the resulting anxiety is not seen as intrinsic or related to the 
personality but (like the diffuse tension or physical symptom) is seen 
as external. 

Turning from the symptomatic picture to the clinical impression 
given by these patients, we see a new set of diagnostic cues. 

The bland overconventional person is often seen as immature by the 
clinician. This term runs through the typical case reports and reflects 
the naivete, the artless, childlike ingenuousness which these security 
operations maintain. These patients see no evil, hear no evil, think no 
evil, do no evil. They handle interpersonal situations by complaisant, 
serene machinery. 

The rigidity by means of which these individuals can distort and 
misinterpret reality can reach astounding proportions. These misper- 
ceptions (sincere and not deliberate) can lead to disastrous misunder- 
standings. We think here of the patient who employed two solid layers 
of bland optimistic friendliness to handle feelings of despair so severe 
as to reach psychotic proportions. In the face of several catastrophic 
failures (loss of two jobs, threatened divorce), this patient insisted in 
the intake interview that everything was going well, that he was not 
depressed, etc. 

The discrepancy between the reality situation and his happy re- 
actions finally emerged. The intake worker reviewed with the patient 
the intense conflict between desperate fearful depression and the 
cheerful fagade. The latter operations were supported, but the need 
for treatment was stressed. The patient was delighted with the course 
of the interview, enthusiastically accepted the mild summary of the 
clinician, and eagerly cooperated in making plans for therapy, arrang- 
ing future appointments, etc. 

Within two days the clinician received phone calls from three irate 
and puzzled people (his wife, his employer, and the referring phy- 
sician), all of whom had been informed by the patient that "the psy- 
chiatrist said I am perfectly normal and don't need treatment." In a 
subsequent interview the patient remembered the negative or reality 
side of the clinician's original summary and stated that he had "for- 
gotten" the plans for therapy and discovered the appointment slip 
which had been "lost" in his wallet. 

This patient was not a dishonest or prevaricating person. The 
rigidity and intensity of the ingenuous naivete, as well as the com- 
plete crippling of any other interpersonal reflexes, were quite evident 


in his Level I and II test patterns and testified eloquently to the pres- 
ence of a blanket denial process which made it intolerable for this 
person to face unpleasant reality. 

This same process is regularly observed in group therapy where 
hysterical patients completely misperceive hostility in others and for- 
get the occasion when they have been momentarily angry or depressed 
in the group. 

The Relationship of Over conventionality to 
Standard Psychiatric Diagnosis 

Patients who manifest the reflexes of intense, maladaptive over- 
conventionality are often given the psychiatric diagnosis of hysterics. 

If we review the clinical characteristics of the conventional per- 
sonality, we will observe that they tend to fit the general conception 
of hysterical behavior — the physical symptoms, bland denial of emo- 
tional problems, etc. 

The chapters in psychiatric texts which describe the hysteric usu- 
ally center the discussion around the dramatic symptomology: fugues, 
amnesias, paralyses, etc. (3,2) In recent years these colorful symptoms 
tend to appear in diminishing frequency and the diagnosis of hysteria 
is increasingly being based on dynamic, interpersonal or psychosexual 
criteria (4, 1). 

The bland, optimistic conventionality of the hysteric has been dis- 
cussed in the literature for over sixty years. Charcot defined this diag- 
nostic characteristic in describing "la belle indifference des hyster- 

The current trend in diagnosis seems to emphasize the dynamic 
aspects of the hysterical personality. Schafer in his competent diag- 
nostic volume (4) consistently employs "functional" or dynamic 
variables rather than symptomatic or descriptive cues. He defines 
hysterics as persons who "rigidly and persuasively resort to the defense 
of repression." He speaks of the narrowed cultural and intellectual 
interests, the impaired ability to think independently or to express 
original, individual themes. He also refers to the naivete of these pa- 

In the interpersonal diagnostic system the term hysterical per- 
sonality is used to describe patients whose presenting operations stress 
bland, narrow conventionality. It must be emphasized that we are 
diagnosing overt operations. Hysterics vary considerably in their un- 
derlying motivation. Some patients employ an optimistic overcoop- 
erative fagade to mask underlying schizoid or sadistic feelings. Other 
overt hysterics present solid, four-layer structures of friendly con- 


When we employ the term hysteric, we do not necessarily desig- 
nate the "simple-hysteric-serving-girl" syndrome for which sugges- 
tion and hypnosis have traditionally been used as therapeutic tools. 
Patients with overt hysterical operations can be very complex in their 
multilevel patterns. A wide variety of treatment regimes can be 
recommended depending on the nature of the underlying material, 
the ability to tolerate the warded-off emotions, etc. 

Research Findings Characteristic of the "77" Personality 

Here is a summary of the empirical studies of the overconventional 

1 . Patients who utilize overconventional security operations present 
MMPI profiles which stress hysteria (Hy) and denial-of-psycho- 
pathology (K) and which underemphasize schizoid isolation (Sc) 
and nonconformity (F). 

2. These patients are not well motivated for psychotherapy. They 
remain in treatment for an average of nine sessions and rank fifth 
among the eight diagnostic types on this variable. 

3. They are closely identified (consciously) with their mothers and 
their marital partners — ranking third among the eight diagnostic types 
on this variable. 

4. They are the most consciously identified with their fathers of 
any diagnostic group. 

5. They misperceive the behavior of others by attributing too 
much friendliness and affiliativeness to others. They tend to blanket 
others (in their therapy groups) with the same conventional sweetness 
that they claim for themselves. 

6. The overconventional personality appears in certain cultural 
and institutional samples more frequently than others. The percentage 
of hysterics in various samples at Level I-M is presented in Table 22. 

The highest percentage of overconventional cases is presented by 
the self-inflicted dermatitis group. These patients bend over back- 
wards to inhibit the unconventional at Level I, although in their fan- 
tasies they are more hostile than any other sample (see Chapter 24). 

Another high percentage of hysterical subjects is found in the hy- 
pertensive example. This is an expected result. It has been repeatedly 
claimed in the psychosomatic literature that hypertensives present un- 
usually sweet friendly fagades. This bit of clinical folklore has been 
confirmed by the Kaiser Foundation research studies on psychoso- 
matic subjects, in which we have found hypertensives presenting the 
facades of conventionality. 

An equally high percentage of hysterical subjects is found in the 
individual therapy sample. We have already noted that they do not 




































Percentage of Cooperative-Overconventional Personalities (Level I-M) 
Found in Several Cultural Samples 

% of Cooperative-Overconventional 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 
College Undergraduates 
University Psychiatric Clinic 
Middle Class Obese Patients (Female) 
Overtly Neurotic Dermatitis Patients 
Self-inflicted Dermatitis Patients 
Unanxious Dermatitis Patients 
Group Psychotherapy Patients 
Individual Psychotherapy Patients 
Hypertensive Patients 
Ulcer Patients 
Medical Control Patients 
University Counseling Center 
University Graduate Students (Male) 
Stockade Prisoners (Male) 
Hospitalized Psychotic Patients 
Officers in Military Service 

Total 1903 

tend to remain long in treatment. The identity of the groups with the 
next highest frequency of hysterical personalities is, however, some- 
what startling. The ulcer sample, the obesity sample, and the stockade 
prisoner sample are tied for third rank. This suggests that the most 
conventional sample (middle class women seen in a nonpsychiatric 
setting) and the most antisocial sample (prisoners) share the same 
percentage of naive, bland subjects! 

From the standpoint of a multilevel theory, the latter result is not 
completely unexpected. Many of the most severely antisocial indi- 
viduals present fa9ades of piety and virtue which are almost painful to 
observe. Alany delinquents, addicts, criminals are characterized by a 
bland, naive, innocent front behind which rage intense feelings of dis- 
trust or rebellion. 

The same is true of some institutionalized psychotics. This fact has 
confused some psychologists who observed that some of the most se- 
verely disturbed paranoids and deluded schizophrenics presented hys- 
terical, repressive MMPI profiles. Some state-hospital psychologists 
have reported that the MMPI is invalid because so many psychotic 
patients have conventional, bland records. This objection completely 
misses the multilevel complexity of personality. Many hospitalized 
patients are psychotic because they tried to maintain a brittle fa9ade of 
pious, self-satisfied virtue and were unable to tolerate their own intense 


hostile or guilty feelings. Many psychotics and delinquents attempt 
to preserve the appearance of naive innocence and are diagnosed (by 
the interpersonal system) as hysterics or hypernormals (at Levels I 
and II) with underlying feelings of a more antisocial nature. 

7. The frequency of this personality type at Level II-C is slightly 
different from the picture at Level I-M (see Table 23). The sample 
of normal controls (labeled medical controls) manifests the largest 
percentage of overconventional personalities. Individual therapy pa- 
tients again provide a larger percentage of this type. The psycho- 
somatic and neurotic samples run about equal to chance expectancy 
except for the ulcer group. The self-diagnosis of this latter sympto- 
matic sample seems to emphasize stronger and more competitive feel- 
ings (see Chapter 24). 


Percentage of Cooperative-Overconventional Personalities (Level II-C) 
Found in Several Cultural Samples 

% of Cooperative-Overconventional 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 
Hospitalized Psychotic Patients (Male) 
Group Psychotherapy Patients 
Individual Psychotherapy Patients 
Overtly Neurotic Dermautis Patients 
Self-inflicted Dermatitis Patients 
Unanxious Dermatitis Patients 
iMedical Control Patients 
Ulcer Patients 
Hypertensive Patients 
Middle Class Obese Patients (Female) 

Total 781 

8. Hysteric patients consciously perceive their parents to be sweet 
and docile. They describe their fathers as being conventional and 
loving (ranked third out of the eight diagnostic groups on the LM 
axis). They see their mothers as being sweet, docile, and trustful 
(ranked second on the affiliative axis and third on the passivity axis). 

9. These patients, on the contrary, see their spouses as relatively 
hostile (ranked fifth on the affihative axis). They describe their 
marital partners as more hostile than do the schizoid and narcissistic 
patients. This seems to fit in with the clinical finding that these pa- 
tients come to the chnic not because of dissatisfaction with their own 
character structure or with their past life (e.g., their parents) but 
because of current external stress (which often involves misbehavior 
of or rejection by their spouses) . The hysterics thus "cross the circle," 
























attributing themes to their marital partners which are the opposite of 
their own self-conceived sweetness. Narcissistic patients, it might be 
noted, do the opposite. They present themselves as superior, snobbish, 
competitive, and somewhat exploitive. They picture their spouses as 
being the most naive, docile, and gullible. 


1. Fenichel, O. The psychoanalytic theory of neurosis. New York: Norton, 1945. 

2. Landis, C, and M. Marjorie Bolles. Text book of abnormal psychology (rev. 
ed.). New York: The Macmillan Co., 1950. 

3. Malamud, VV. The psychoneuroses. In J. McV. Hunt (Ed.), Personality and the 
behavior disorders. New York: The Ronald Press Co., 1944. 

4. ScHAFER, Roy. The clinical application of psychological tests: Diagnostic sum- 
maries and case studies. New York: International University Press, 1948, p. 34<5. 


Adjustment Through Responsibility: 
The Hypernormal Personality 

In this chapter we shall become acquainted with the responsible hyper- 
normal personality type. This comprises those patients whose overt 
behavior locates in the ON octant of the diagnostic grid. This is the 
"88" personality type. These individuals employ strong and conven- 
tional security operations. They present themselves as reasonable, 
successful, sympathetic, mature. They avoid the appearance of weak- 
ness or unconventionality. 

Adaptive Forms of the Responsible Personality Type 

Here we deal with the individual who attempts to present himself 
as a "normal" person. He presents himself as strong — but his power 
and self-confident independence are used in an affiliative way. He 
strives to be close to others — to help, counsel, support, and sympathize. 
He wants to be seen as tender with his intimates, reasonable and re- 
sponsible with his acquaintances. 

These individuals often give the impression of maturity and parental 
strength. They appear sound, sympathetic, considerate. They are 
often popular figures — they attempt to get along well with others 
and to provoke admiration from others. They are leaned upon and 
depended upon by other people. They strive to fulfill an idealized 
role of successful conventionality. 

Maladaptive Forms of the Hypernormal Personality 

An inflexible, repetitious use of responsible, hypernormal reflexes 
leads to a maladaptive condition. Individuals of this type cannot take 
a passive or aggressive or bitter role, even when it is called for. They 
avoid these latter behaviors so compulsively that they become carica- 
tures of hypernormality. 



These individuals "knock themselves out" to be popular. Their 
attempts to be helpful and responsible are often inappropriate. They 
may overextend themselves in promises to others — offers of help and 
sympathy which they cannot fulfill. They may desperately attempt 
to maintain the fa9ade of normality when the situation and their own 
private feelings involve other reactions. They are often driven by 
relentless ideals of service and contribution to others. 

Extreme NO behavior inevitably leads to a bland, uninsightful 
fagade. These individuals cannot tolerate unconventional or weak 
feelings. They are so compulsively attached to their hypemormal 
strivings that they completely deny and inhibit feelings of frustra- 
tion and passivity. 

These extreme operations generally indicate severe conflicts which 
are expressed not in the classic symptoms of neurosis, but in indirect 
(often psychosomatic) manifestations. 

The Purpose of Responsible or Hypemormal Behavior 

Those human beings who are strong and conventionally normal 
have selected these operations because they find them most effective 
in warding off anxiety. Their feeUngs of self-esteem are bolstered by 
appearing mature and generous. They are most secure when they are 
involved in close, friendly protective relationship with dependent 

They are, we assume, most threatened by the prospect of appear- 
ing defeated, deprived, unfriendly, or passive. Their genial, generous 
operations tend to relieve feelings of anger, helplessness, or isolation. 
They give the assurance (at least consciously) of being involved in 
tender, protective relations with others. 

There are many obvious rewards to the responsible hypemormal 
way of life. This mode of adjustment is close to the cultural ideal. 
It thus brings great conscious superego satisfactions. These patients 
are the most self-satisfied individuals seen in the clinic. 

In its adaptive form this is a most positive and socially constructive 
personality type. In that Utopian society where skepticism, sternness, 
competition, or modesty would not be necessary responses and where 
consistent aflSliative behavior would be appropriate, the generous NO 
type would be the rule. Even in the nonutopian twentieth century 
culture the ideals of tender, protective nurturance are undoubtedly 
the most appealing standards. The conventionally successful and 
popular person in our society is usually the one who employs the NO 
interpersonal reflexes a large part of the time. 

Compulsive and inappropriate maintenance of these operations 
leads to the phenomenon of the "hollow man" — isolated by his self- 


satisfied piety from the realities of life and (more dangerously) from 
his own inner feelings of bitterness or weakness. 

The Effect of "88" Behavior 

Responsible, protective behavior pulls dependence and respect 
from others. The person who overtly gives tends to attract those who 
want to receive. In the language of the interpersonal system, NO 
pulls KL. 

Tender, supportive operations tend to train others to agree, con- 
ciliate, and depend. This rule (like the previous generalizations about 
interpersonal reciprocity) is a probability statement. Generosity does 
not always pull friendly dependence. Those who are rigidly com- 
mitted to other interpersonal reflexes will react to the reasonable, gen- 
erous person with their favored responses. In general, however, most 
people tend to expect good things from those who promise good 

Another exception to this rule of reciprocity (i.e., NO pulls KL) 
occurs when the hypernormal behavior is extreme or inappropriate. 
Here we think of the overmotherly woman, the compulsive popularity 
seeker, the overprotective parent. While these behaviors generally 
tend to pull cooperative dependence, their uncalled-for intensity may 
eventually provoke resentment or frustration from the "other one." 

In therapy groups, the "88" individual takes the role of the assistant 
therapist. He encourages, suggests, and sympathizes with the other 
patients. He does not exhibit needy or helpless reflexes but is seen as 
the competent helpful leader. 

Typically the group members cannot understand why he is in 
therapy. They see his reasonable, generous fagade as an ideal adjust- 
ment. At this point the interpersonal network tightens. The other 
patients increasingly put more dependent pressure on the "88" person. 
He is now being asked for help, expected to give, and is given less and 
less allowance to present his own problems as a fellow patient. 

These patients are thus the popularity leaders of the group. By 
acting in a hypernormal way they are not seen as patients needing 
help. They build up an interpersonal process which would lead to 
their getting no therapeutic help from group therapy. At this junc- 
ture the task of the therapist is to step in and assist the "88" person in 
understanding what he has done to the others to block himself off 
from the possibility of help. 

Clinical Manifestations of the Hypernormal Personality Type 

Patients whose overt security operations strive towards normality 
do not present the typical neurotic symptoms when they appear in 


the psychiatric clinic. They are not anxious or depressed. They do 
not report interpersonal failures. They do not complain of timidity, 
isolation, distrust, etc. They tend to describe their emotional adjust- 
ment as adequate and normal. 

Why then, do they come to the clinic? In the Kaiser Foundation 
clinic which services a large general hospital, 23 per cent of all appli- 
cations are hypernormal individuals. The overwhelming majority of 
these patients are not self-referred, but have come at the request of a 
physician. Their symptoms are psychosomatic or physical. 

From 40 to 50 per cent of patients with psychosomatic diseases fall 
in the NO octant. (The frequency expected by chance is 12.5 per 
cent.) Seventy-nine per cent of psychosomatic patients fall in the NO 
octant or its two neighboring octants. For this reason patients who 
locate in this sector of the diagnostic grid can be called psychosomatic- 
type personalities. 

It must be pointed out that we refer here to organ neurosis condi- 
tions and not to somatic expressions of anxiety (nervous stomach, 
transient pains, etc.). The latter are typical of the docile phobic per- 
sonality. The symptoms manifested by the "88" personality are not 
transitory expressions of tension. The hypernormal personality is usu- 
ally successful in warding off anxiety and presents a bland, strong 

In addition to psychosomatic symptoms the "88" personality often 
comes to the clinic for the purpose of putting indirect pressure on 
family members. They may present a story of marital difficulty in 
which it becomes clear that the spouse or a child is "sick and in need 
of help." The subject may recount a history of patient tolerance of 
the family member — the implication being that the diagnostician will 
give the patient a clean bill of health and suggest that the errant spouse 
be brought in to treatment. The poised, "mature" reasonableness of 
the "88" fagade may tempt the inexperienced clinician into collaborat- 
ing in the plans to inveigle family members into therapy. 

A third reason for the "88" personality coming to the clinic in- 
volves certain forms of isolated behavior disorders such as alcoholism, 
gambling, or certain sexual aberrations. These patients may be self- 
referred or sent in by family pressure or court order. They readily see 
the symptomatic behavior as ego-alien — but isolate it from their per- 
sonality. The motto for these patients might be: "I am a well-adjusted 
nice guy — if only I could get rid of that crazy behavior pattern." 

Many alcoholics or addicts fall into other diagnostic categories. 
The guilty or the defiant types do not, of course, locate in the "88" 
sector of the diagnostic grid. Alany behavior-disorder patients, how- 
ever, do attempt to maintain a repressive hypernormal fa9ade. 


Many cases of impotency or frigidity fall in the hypernormal sec- 
tor. Here again the symptom (like that of the psychosomatic) is seen 
as isolated from the well-adjusted personality. 

Many severely deluded paranoid patients present themselves as 
hypernormal at the symptomatic level. This is really not a paradoxical 
situation if the theory of levels is kept in mind. 

Level I summarizes the patient's impact on the clinician. The es- 
sence of certain forms of many severe paranoid conditions is that the 
patient strives to appear hypernormal. When we assign this Level I 
diagnosis, we do not assume that this means the patient really is 
normal, but rather that his security operations at this level strive to 
create this impression. 

Very often patients reporting to an out-patient clinic after hospital- 
ization for a psychotic break present as hypernormal at the level of 
symptoms. These patients have utilized repressive measures to handle 
their psychotic impulses. They are sitting on their conflicts and striv- 
ing to maintain a fa9ade of conventional strength. The multilevel pat- 
tern and the clinical interview will usually indicate how precarious or 
brittle these surface operations are. 

In summary it can be said that whatever the reason bringing the 
hypernormal patient to the clinic, it is seen by him as an annoying ap- 
pendage separate from his perception of his own personality. This 
situation makes the "88" patient a particularly tricky prognostic prob- 

Standard Psychiatric (Kraepelinian) Equivalents 
of the Hypernormal Personality 

There appears to be no standard psychiatric diagnosis which covers 
the behavior described in this chapter. Psychiatric literature has tra- 
ditionally neglected the normal, the superior, and the supernormal 
personality, and those who present these operations. 

Before the increasing popularity of psychosomatic concepts, the 
"88" personality type did not appear in the psychiatric consulting 
rooms. The early psychiatric theories and nosologies were clinical in 
origin. The overtly strong, popular, protective personality failed to 
receive conceptual attention. 

The Kaiser Foundation clinic (because of its consultative relation 
to a general hospital) has evaluated hundreds of patients whose overt 
and conscious behavior is hypernormal. The Foundation's research in 
psychosomatic medicine has collected multilevel test batteries on more 
than one thousand of these cases. 

Analysis of these protocols has led us to view strong, affihative, 
supportive security operations not as ideal or normal ways of behaving 


but as machinery for warding off anxiety, avoiding disapproval, and 
raising self-esteem. There are several hundred multilevel patterns of 
behavior which can underly a hypernormal fagade. Some of these 
patients might be given psychiatric or psychosomatic labels (depend- 
ing on their specific symptomology). Some of them are psychotic 
individuals (usually paranoid) who desperately cling to an overt ap- 
pearance of adjustment. Many of them would remain undiagnosed 
according to current psychiatric nosology and, if labeled at all, would 
be called "normals." 

Research Findings Characteristic of the Hypernormal Personality 

The Kaiser Foundation research has studied several hundred sub- 
jects with the Level I diagnosis of responsible-hypernormal personal- 
ity. Here is a summary of current findings. 

1. Forty-three per cent of all patients with psychosomatic symp- 
toms fall in the NO octant at Level L Psychosomatic patients there- 
fore use these hypernormal operations three or four times more than 
chance expectancy. 

2. Fourteen per cent of nonpsychosomatic psychiatric patients fall 
in this sector. These patients, therefore, do not use these operations 
more than chance expectancy. 

3. Hypernormal operations are characterized by the following 
MMPI pattern: high scores on hypochondriasis (Hs), repressive 
blandness (Hy), denial of pathology (K), low scores on nonconform- 
ity (F), depression (D), schizoid (Sc), and obsessive tendencies (Ft). 

4. These patients are not well motivated for psychotherapy. They 
rank as the lowest group in average number of therapy sessions. This 
indicates that they refuse treatment or quit soon after beginning. 
Pure or stable hypemormals remain in treatment on the average of six 
sessions. Conflicted hypernormals remain in treatment about twice as 
long (average equals eleven sessions). Here the underlying trends (of 
weakness or bitterness) make them more likely to remain in therapy. 

5. They are highly identified with their parents. In our study of 
conscious identification with mother, father, and spouse, they rank 
first: and on another study, second among the eight diagnostic groups. 

6. Hypernormal patients (along with hysterics) tend to misper- 
ceive the interpersonal behavior of others in a consistent direction. 
They attribute too much friendliness and cooperativeness to others. 
They also tend to see others as stronger than they are consensually 
judged to be. This indicates that responsible personalities consistently 
tend to see others as like themselves — falsely perceiving others as more 
loving and strong than they are. This is unquestionably a function 
of their attempt to maintain a bland, conventional atmosphere which 


fails to take into account the actual amount of hostility- weakness pres- 
ent in others as well as themselves. 

7. The hypernormal personality type appears in certain cultural 
and institutional samples much more frequently than in others. The 
percentage of responsible individuals in various samples at Level I-M 
is presented in Table 24. 


Percentage of RESPONSlBLE-HYPER^"ORM^L Personalities (Level I-M) 
Found in Several Cultural Samples 

7o of Respoiisible-Hypernormal 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 
College Undergraduates 
University Psychiatric Clinic 
Middle Class Obese Patients (Fem.ilc) 
Overtly Neurotic Dermatitis Patients 
Self-inflicted Dermatitis Patients 
Unanxious Dermatitis Patients 
Group Psychotherapy Patients 
Individual Psychotherapy Patients 
Hypertensive Patients 
Ulcer Patients 
Medical Control Patients 
University Counseling Center 
University Graduate Students (Male) 
Stockade Prisoners (Male) 
Hospitalized Psychotic Pauents 
OflScers in Military Service 

Total 1903 

The percentage figures expected by chance for these groups is 12.5. 
The psychiatric sample contains more than the expected percentage 
because the Kaiser Foundation clinic services a general hospital and 
four additional medical centers. The number of patients with somatic 
and psychosomatic referrals is much greater than that seen in the 
standard psychiatric clinic. The clinic policy of referring certain 
naive, conventional, or hypernormal patients to group therapy for 
educational reasons accounts for the fact that a higher percentage of 
responsible patients are seen in group therapy. 

In general it will be noted that the noncHnic samples of "normal" 
subjects (e.g., army officers) contain three to four times the expected 
percentage of hypernormal subjects. 

8. The percentage of responsible-hypernormal individuals in var- 
ious samples at Level II-C is presented in Table 25. It will be observed 
that the psychosomatic samples tend to have three times the expected 
number of hypernormal patients. The ulcer sample and the neuro- 


























dermatitis samples, whose respective tendencies towards aggression 
and masochism have been previously noted, are exceptions to this 
generalization. The fact that a fairly high percentage of psychotics 
claim to be hypernormal is an interesting finding, the significance of 
which is discussed in Chapter 23. 


Percentage of Responsible-Hypernormal Personalities (Level II-C) 
Found in Several Cultural Samples 

% of Responsible-Hypernormal 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 
Hospitalized Psychotic Patients (Male) 
Group Psychotherapy Patients 
Individual Psychotherapy Patients 
Overtly Neurotic Dermatitis Patients 
Self-inflicted Dermatitis Patients 
Unanxious Dermatitis Patients 
Medical Control Patients 
Ulcer Patients 
Hypertensive Patients 
Middle Class Obese Patients (Female) 

Total 781 

9. Hypernormal patients see their fathers as exceedingly strong- 
conventional people. The father is consciously idealized. The mother 
is seen as extremely loving, tender, and agreeable. A most conven- 
tional portrait of both parents is produced. They also see their marital 
partners as conventional, friendly — but slightly more docile than their 
conscious picture of their mothers. The conscious descriptions of all 
three family members are located on the conventional side of the 
diagnostic grid. 

10. Of all the "88" patients seen in the psychiatric clinic over a one 
year period, 74 per cent did not go into psychotherapy. By compari- 
son only 46 and 48 per cent of distrustful and masochistic patients did 
not go into therapy. This lends empirical support to the statement 
that hypernormal subjects are not initially well motivated for psycho- 
























Adjustment Through Power: 
The Autocratic Personality 

Power, success, and ambition as means of warding off anxiety and in- 
creasing self-esteem comprise the theme of this chapter. We shall con- 
sider those individuals whose overt interpersonal operations stress 
compulsive energy, authority, and dominance over others. This is the 
"11" personality type. 

Until recently, these patients were not often seen in psychiatric 
consulting rooms. The nature of their security operations is such that 
they were not seen as needing psychiatric help and would hardly con- 
sider asking for help. 

Adaptive Forms of the Power-Oriented Personality 

Adjustment through power can be an adaptive and successful way 
of life. Included here are those persons who express strength, force, 
energy, and leadership, and who win from others respect, approbation, 
and deference. 

The generic idea of hero belongs to this mode of adjustment. So 
do all forms of ambition. So do the traits of energy, planful organiza- 
tion, and righteous authority. Behavior which is designed to excite 
admiration or to provoke submission from others can be considered 
as diagnostic of this security operation. 

There are many ways in which power can be manifested. Physical 
strength, especially in the case of the male, is a means of winning 
respect. Intellectual strength is another common power operation. 
The sage, the wise man, and the savant are all roles which earn respect. 
The interpersonal mechanism of teaching is, in fact, probably the most 
common manifestation of power motivation. The ordinary, common- 
place frequency of the teaching behavior makes its power implications 
go unnoticed. It seems clear, however, that whenever one person be- 




gins to instruct, inform, or explain to another, he is conveying the 
interpersonal message, "I know something you do not know, I am 
wise and better informed on this subject than you." Intellectuals are 
often power-oriented individuals who maintain illusions of strength 
and prestige through their knowledge. The nonintellectual who can- 
not understand why teachers seek out and remain in positions of such 
low pay may fail to recognize the rewards and securities which accrue 
to the pure undiluted power expression of the pedagogue. 

Teaching is thus a most adaptive and constructive manifestation of 
the autocratic impulse. 

In addition to physical and intellectual strength, there are several 
other ways in which power can be gained and expressed. Social 
status is perhaps one of the most effective means of exerting authority. 
Prestige — either bureaucratic or social — is a power magnet for at- 
tracting respect and deference. 

Financial strength is another common form of power expression. 
Most forms of conspicuous consumption are diagnostic of the attempt 
to maintain a superior (22) or powerful (11) fa9ade. 

In summary it can be said that the "11" personality is characterized 
by energetic, organized behavior, by the attitude of knowledge, com- 
petence, strength, and authority. 

Maladaptive Forms of the Poiver-Oriented Personality 

The extreme forms of this way of life are characterized by auto- 
cratic, domineering behavior. Compulsive attempts to control are 
diagnostic of this maladjustment. So is power-ridden, overambitious 
behavior. Pedantry falls into this category; as do status-driven at- 
tempts to impress. 

The person who tries to overorganize his life and the lives of those 
around him is utilizing maladaptive power operations. The compulsive 
person is often striving to increase his fa9ade of competence and ef- 
ficiency. His exaggerated attempts to be planful, precise, and correct 
are diagnostic of the "11" maladjustment. 

The key factor in this maladaptive type is the complete avoidance 
of weakness and uncertainty, and the compulsive endeavor to appear 
competent, organized, and authoritative. 

The autocratic person exhibits his power-oriented machinery of 
adjustment rigidly whether it is appropriate to tne situation or not. 
He cannot relax his compulsive, energetic operations. In social or 
recreational contexts he grimly clings to his mantle of efficiency and 
competence however uncalled-for it may be. The extremes of thi^ 
type of maladjustment often involve hyperactivity and manic be- 


In the clinic the autocratic individual is thus easily diagnosed by his 
inappropriate responses. He does not act like a patient coming for 
help — but as a strong competent person seeking to inform or impress 
the clinician. 

The Purpose of "11" Behavior 

Individuals select power-oriented security operations because they 
have found them to be effective in warding off anxiety. They feel 
secure when they are exerting control over people and things. They 
apparently dread the possibility of being weak, uninformed, submis- 

The rewards and comforts which can be obtained through control 
and power are numerous. The strong person feels defended and pro- 
tected. He wins awe, admiration, and obedience from others. He 
gains a feeling of certitude and organization — which serve as an il- 
lusory buffer against the mysteries and uncontrollable possibilities of 

The autocratic individual is, we assume, made most anxious when 
he feels uncertain, confused, or passive. He attempts to maintain 
security and self-esteem and to avoid derogation and hurt by means of 
his power-oriented operations. 

The Effect of "11" Behavior 

The fagade of power and control provokes others to obedience, 
deference, and respect from others. This is to say, AP pulls //. 

In most situations the person who manifests wisdom is looked to 
for advice. The person who demonstrates planful control and compe- 
tence is respected. 

This principle of reciprocal interpersonal relations is, of course, a 
probability statement. It can be altered by the personality of the other 
person. Thus a managerial person interacting with another who uses 
the same interpersonal reflexes may generate a power struggle. He 
may receive agreeable cooperation from a person with hysterical op- 

In general it will be found that rigid autocratic individuals seek out 
docile admiring followers. They are most comfortable when they 
are paired with those who symbiotically match their interpersonal 
reflexes — who flatter, obey, and respect them. 

Clinical Manifestations of Managerial Power 

It has been pointed out that prior to the 1930's the managerial per- 
sonality was not a frequent visitor to the psychiatric clinic. In recent 
years, however, a broader definition of neurosis (as any form of ex- 


treme or maladaptive behavior) has developed. In addition, the con- 
cepts of psychosomatic medicine have stressed the point that certain 
physical symptoms can be manifestations of maladaptive conflicts. 

For these reasons, more and more patients whose overt fagade 
stresses power and energy are being referred for psychiatric diagnosis. 
There are several specific clinical characteristics of the power-oriented 

Psychosomatic symptoms are a most common complaint. Ulcer 
patients are classically seen as driving, ambitious, energetic people. 
Certain dermatitis diagnostic groups utilize strong interpersonal re- 
flexes. Overweight women tend to present clinically in the same man- 
ner. Asthmatic men tend to stress power and deny weakness in their 
approach to a psychiatric clinic. 

Some strong managerial individuals come to the clinic because of 
their concern about other family members. One frequent type of 
referral involves the competent, industrious woman who is married to 
a weak, delinquent, or rebellious husband. The managerial wife comes 
partially seeking the clinic's support in getting her husband into treat- 
ment and partially because of her own underlying passive needs. This 
type of strong woman inevitably manifests "preconscious" masochism 
and is usually involved in a complicated guilt-power conflict with her 

Many cases of alcoholism or gambling present a power-oriented 
facade to the clinic. These patients see their symptomatic behavior as 
isolated from their character structure and are not initially well 
motivated for therapy. The prognosis in these cases depends upon 
the ability to tolerate consideration of their underlying rebellious or 
passive feelings. 

Another symptom typical of the "11" personality involves an iso- 
lated anxiety attack. The patient regularly uses compulsive, energetic, 
self-confident operations to handle anxiety. This fagade may tem- 
porarily crack (in response to a particularly threatening environmental 
circumstance). The patient comes to the clinic because he is scared 
by the possibility of a recurrence. (An anxiety attack or any other 
sign of weakness is, of course, the most paralyzing catastrophe to the 
person who utilizes power security operations.) By the time the 
patient comes for his intake interview, his routine compulsive reflexes 
may be working smoothly again. He mobilizes against the threat of 
anxiety created by psychological exploration and presents a fa9ade of 
competent strength. These patients see their anxiety attacks as iso- 
lated events, not integral to their strong character structure. The lat- 
ter they do not usually want to change. 


Some managerial personalities (male) come to the clinic with symp- 
toms of impotency. The fagade of strength is particularly disturbed 
by sexual inadequacy. Generally these patients are eager to have the 
symptoms (which are uncomfortable signs of weakness) removed and 
are not pressing to explore the underlying passivity or fear which the 
symptoms represent. 

(Occasionally some "11" types come to the clinic because of dis- 
satisfaction with their interpersonal relationships with others. The 
competent wife puzzled by her errant husband has been mentioned. 
The compulsive, righteous husband frustrated by a rebellious wife, or 
by resentful children, is another example. Now and then compulsive 
patients come under pressure from their employers who threaten to 
fire them because of friction generated by their power strivings. A 
particularly sad variety of managerial operations is afforded by the 
masculine, driving woman who finds herself lonely and neglected by 
men and who hopes to find relief from her vague dissatisfaction with 
self without relinquishing her compulsive protections. 

There is one exception to this generalization. Some highly intelli- 
gent, psychologically sophisticated individuals come to the clinic seek- 
ing intensive treatment or psychoanalysis. These patients are actu- 
ally hoping to change their character structure. They may have some 
of the symptoms mentioned above and are insightful enough to want 
therapeutic help. These patients are intellectually (and not emotion- 
ally) motivated for psychotherapy. They will exhibit their power re- 
flexes but have enough insight to ask for and remain in treatment. 
Such patients are usually referred to psychoanalysts or assigned to 
intensive psychotherapy. 

Relatio?2ship of Power-Oriented Personality 
to Standard Psychiatric Diagnostic Types 

Although adjustment (or maladjustment) through power has not 
classically been the focus of much psychiatric theory, there are two 
diagnostic types which have some of these interpersonal factors im- 
plicit in their definition. 

The compulsive personality seems to involve definite power mo- 
tives. The compulsive person is one who is active, prompt, well-or- 
ganized, industrious, pedantic, planful, and often righteously compe- 
tent. The person who exhibits these traits is clearly trying to impress 
others with his effectiveness. (The fact that he is generally trying to 
deal with his own inner feelings of guilt or impotency may appear in 
the form of multilevel conflicts which often characterize the com- 
pulsive patients seen in the clinic.) 


Successful, well-adjusted compulsives are generally respected by 
others for their diligence and organization. The notion of efficiency 
(for the American and German cultures, at least) is heavily loaded 
M^ith power connotations. In Chapter 16 we have attempted to dis- 
tinguish between the obsessive and the compulsive modes of adjust- 
ment. It was suggested there that these two behaviors are quite dis- 
tinct in terms of symptom and interpersonal meaning to others. 

The obsessive person usually presents as guilty, passive, and un- 
certain. The compulsive as strong and right. While their security op- 
erations are different, they can sometimes be seen in the same per- 
sonality pattern, usually when the compulsive defenses are breaking 

The specific power elements of the compulsive state have not been 
made the central diagnostic key — usually being subordinated to 
symptomatic factors. Some of the dominance-submission aspects of 
compulsivity are implicit in certain psychoanalytic writings. Freud's 
first and most authoritative paper on the compulsive character was 
published in 1908 (2). At this time he presented his conception of 
the three anal characteristics: orderliness, parsimony, and obstinacy. 
In the paper "Character and Anal Eroticism" he first described the 
first great power struggle of life: children's "great self-will about 
paning with their stools." He then describes the parents typical at- 
tempt to "break his (the child's) self-will and make him submissive." 
Fenichel (1, p. 280) sees the anal character trait of orderliness as "the 
elaboration of obedience." 

Most psychoanalytic writers tend to agree as to the power elements 
of the compulsive personality but draw psychosexual rather than inter- 
personal conclusions. Compulsivity is thus seen as a reaction forma- 
tion against the child's stubborn, managerial wish to foil the parent by 

Mullahy (3, p. 61) has presented a summary of the resolution of 
this archaic power struggle which is very congenial to the inter- 
personal theory. He points out the strivings for "self-determination" 
associated with anal activities and then makes the additional (and 
crucial remark) : "When the child succeeds in making a virtue out of 
necessity, he is said to identify himself with the requirements of his 
educators and is proud of his attainment. Thus, the primary injury to 
his narcissism is compensated, and the original feeling of self-satisfac- 
tion in being 'good.' " 

Compulsivity (through identification) thus provides the individual 
with the feehng of power and righteousness. 

This relation between self-satisfaction and power is confirmed by 
the empirical findings of the Kaiser Foundation research. Managerial 


personalities are most closely identified (consciously) with their par- 
ents. They are closely identified with their ego ideals. They are 
pleased with themselves. 

Obsessive patients, on the contrary, are the most self-disapproving 
and are least identified consciously with their ego ideals. 

A clinical description of a personality type which seems similar to 
the managerial personality has been presented by Frank et al. (4, p. 
215) They entitle this behavior pattern the doctor'' s assistant, which 
they say, ". . . consists of a patient's tendency in the group to de- 
fend authority, to please the doctor, to offer advice to other patients, 
to hide his own weaknesses, and generally to impress everyone with 
his own excellence. This behavior springs from an idealization of 
authority in general and a conviction that the way to win an author- 
ity's good will is to demonstrate one's loyalty and excellence." 

Research Findings Characteristic of the Managerial Personality 

The Kaiser Foundation research has studied over 2,000 psychiatric 
clinic patients and over 1,000 psychosomatic and normal subjects. 
The managerial type (at Level I) comprises the largest percentage of 
cases studied. Over 600 patients who employ these security operations 
have been diagnosed. We shall now consider some of the current 
research findings characteristic of this personality type. 

1. Twenty-three per cent of patients with psychosomatic symp- 
toms (i.e., ulcer or hypertensive) fall into the managerial sector at 
Level L Only 8 to 13 per cent of nonpsychosomatic patients (i.e., pa- 
tients with classic neurotic symptomatology) are given the diagnosis of 
power-oriented personality. This tends to confirm the suggestion that 
managerial patients do not tend to come to the clinic or enter psycho- 
therapy with overt psychopathological symptoms. 

2. Patients who consciously describe themselves as managerial have 
a characteristic MMPI profile. They manifest high scores on the 
hyperactivity scale (Ma) and the denial of symptoms scale (K), and 
low scores on depression (D) and obsessive tendencies (Ft). They do 
not stress emotional symptoms. 

3. Managerial patients do not tend to enter or remain in psycho- 
therapy. They are seen in treatment on the average of six sessions. 
They rank lowest (tied with hypernormal and narcissists) on number 
of times seen in the psychiatric clinic. They are, therefore, not initially 
well motivated for psychotherapy. 

4. Managerial patients tend to be closely identified (consciously) 
with their parents. On one study they rank first in closeness of identi- 
fication with mother; and on a second study they rank second on this 
variable (being topped only by hypernormals). 



5. They are closely identified with their marital partners. They 
rank second in this variable. Only the hypernormal group claims a 
closer connection with their spouses. 

6. Managerial patients (along with narcissists) have a characteristic 
misperception of the interpersonal behavior of others. They attribute 
too much weakness to others with whom they interact. They seem to 
look down on others and fail to perceive strength in others. 

7. The managerial personality tends to appear in certain cultural 
and institutional settings with varying frequencies. Table 26 presents 
the percentage figures for Level I-M. The percentage expected by 


Percentage of Managerial-Autocratic Personalities (Level I-M) 

Found in Several Cultural Samples 

% of Mavagerml-Aiitocratic 

Institutional or Symptomatic Sample 



Psychiatric Clinic Admissions 



College Undergraduates 



University Psychiatric Clinic 



Middle Class Obese Patients (Female) 



Overtly Neurotic Dermatitis Patients 



Self-inflicted Dermatitis Patients 



Unanxious Dermatitis Patients 



Group Psychotherapy Patients 



Individual Psychotherapy Patients 



Hypertensive Patients 



Ulcer Patients 



Medical Control Patients 



University Counsebng Center 



University Graduate Students (Male) 



Stockade Prisoners (Male) 



Hospitalized Psychotic Paaents 



Officers in Military Service 





chance for these groups is 12.5. It will be noted that all groups except 
the two psychotherapy samples contain much more than the expected 
frequency of managerial personalities. The norms on which these in- 
terpersonal diagnoses are based were taken from a sample of 807 
psychiatric clinic admissions. The results listed in Table 26 suggest 
that more than three times as many military officers manifest power- 
oriented operations than do clinic admission patients. Fifteen per 
cent of admissions to the Kaiser Foundation psychiatric clinic exert 
strong compulsive symptomatic pressure on the clinic. Only 8 per 
cent of individual therapy patients utilize these overt operations, which 
tends to confirm the statement that managerial compulsive patients 
come to the clinic for diagnosis (and perhaps symptomatic relief) but 


do not tend to enter psychotherapy. The larger percentage ( 1 3 per 
cent) of managerial patients in group therapy is caused by the chnic's 
policy of placing certain psychosomatic patients (e.g., ulcer patients) 
in group treatment. 

8. The frequency figures for the occurrence of the managerial per- 
sonality at Level II-C are presented in Table 27. Three psychosomatic 
groups (ulcer, hypertensive, and obese) claim to be stronger by a ratio 
of over 4 to 1 than normal controls. A fairly large percentage of 
psychotic patients attempt to maintain the conscious illusion of execu- 
tive power. Patients who end up in individual psychotherapy are, as 
noted before, docile and less managerial in their fagade operations. 


Percentage of Managerial-Autocratic Personalities (Level II-C) 
Found in Several Cultural Samples 

% of Managerial-Autocratic 
Institutional or Symptomatic Sample N Personalities 

Psychiatric Clinic Admissions 207 12 

Hospitalized Psychotic Patients (Male) 46 17 

Group Psychotherapy Patients 101 12 

Individual Psychotherapy Patients 38 8 

Overtly Neurotic Dermatitis Patients 31 10 

Self -Inflicted Dermatitis Patients 56 13 

Unanxious Dermatitis Patients 70 11 

Medical Control Patients 41 7 

Ulcer Patients 42 33 

Hypertensive Patients 49 33 

Middle Class Obese Patients (Female) 100 32 

Total 781 

9. Managerial patients tend to see their mothers as exceedingly 
strong, independent people. They also describe their fathers as strong, 
but not as powerful as their mothers. They describe their marital part- 
ners as much more passive and agreeable than their parents. 

10. Sixty-eight per cent of all managerial patients seen in the psychi- 
atric clinic (over a one-year period) did not go into treatment. This 
indicates that this personality type is not initially well motivated for 
psychotherapy. (By contrast 46 per cent of distrustful patients did 
not go into therapy.) 


1. Fenichel, O. The psychoanalytic theory of neurosis. Nev/ York: Norton, 1945. 

2. Freud, S. Character and anal eroticism. Collected papers. Vol. 2. London: Ho- 
garth Press, 1948. 

3. Mullahy, p. Oedipus myth and complex. New York: Hermitage Press, 1948. 

4. Rosenthal, D., J. Frank, and C. Nash. The self-righteous moralist in early meet- 
ings of therapeutic groups. Psychiat., 1954, 11, No. 3, 215-23. 


Adjustment Through Competition: 
The Narcissistic Personality 

In this chapter we shall consider a way of life which is based on com- 
petitive self-confident narcissism. This is the "22" mode of adjust- 
ment. This personality type is of particular interest because it ap- 
pears very rarely in the psychiatric clinic and has been given scant 
theoretical attention in proportion to the frequency of its occurrence. 

The "22" personality expresses at Level I a clear love and approval 
of himself. He acts in a strong, arrogant manner. He communicates 
the message that he feels superior to the "other one." He appears in- 
dependent and confident. 

In its adaptive intensity this interpersonal reflex is a most impres- 
sive social maneuver. In its maladaptive extreme it becomes a smug, 
cold, selfish, exploitive social role. In this case the adaptive self-confi- 
dence and independence become exaggerated into a self-oriented rejec- 
tion of others. The individual is so rigidly tied to his own self-enhance- 
ment that he fails to sense the inappropriateness of his behavior. 

Exhibitionism and proud self-display are often diagnostic of this 
personality type. This competitive attitude may show itself in dress, 
carriage, and gesture, or in the purposive meaning of verbalizations. 
The kind of narcissistic expression varies from person to person. Some 
narcissists stress their intellectual superiority. Others (more typically 
women) center their overt narcissism on their appearance, dress, and 
physical beauty. Conspicuous consumption is generally related to 
this security operation as well as all forms of snobbishness. 

The Purpose of Competitive Narcissism 

Narcissistic displays of superiority are a means of warding off anx- 
iety through ascendance and self-enhancement. These individuals 
feel most secure when they are independent of other people and feel 
they are triumphing over them. 



These individuals depend for their self-esteem on the demonstra- 
tion of weakness in others and competitive strength in themselves. 
This security operation is close to the managerial-autocratic. The 
difference lies in the amount of positive or affiliative affect involved. 
The executive personality vi^ants loving respect and obedience. The 
narcissist provokes defeated envy and inferiority feelings. The nar- 
cissist puts more distance between himself and others — he wants to 
be independent of and superior to the "other one." Dependence is 

Another familiar variety of the competitive mode of adjustment is 
seen in the case of the status-driven person. These individuals are un- 
usually sensitive to issues of superiority-inferiority. They invest con- 
siderable energy in protecting and increasing their prestige. This con- 
cern with status is also typical of the managerial personality. The 
latter, however, tends to provoke others to yield authority to him 
willingly through recognition of his strength. The competitive person 
is generally more ruthless and exploitive in his attempts to seize and 
maintain superiority and pulls a less willing submission from others. 
To put it in different terms — the managerial person trains others to 
identify with his strength and gives the impression that his power will 
be used either neutrally or to help the weaker. The narcissistic person 
tends to emphasize his superior difference from the "other" and gives 
the impression that his status and strength will be used to shame or hu- 
miliate the "other." 

Competitive persons are apparently made most anxious by the threat 
of weakness or dependence. Weakness is generally sensed by these 
individuals to be a dangerous or humiliating position. Often the experi- 
ences of childhood have been so traumatic as to lead to a counteraction 
in the direction of strength. In other cases dependence is associated 
with crucial figures with whom the subject desires to disidentify. 
Thus the counterphobic man equates docility with passivity. The 
competitive woman may attempt to act the opposite of a submissive 
parent whose passivity is consciously perceived as a negative trait. 

The specific purpose of independent arrogance is to establish a 
superior invidious relation with others. These subjects apparently 
view passivity, cooperation, trust, or tenderness as dangerous. They 
seem to fear the loss of proud individuality which is attached to these 
other operations. 

There are, of course, many rewards associated with self-confident 
narcissism. Self-approval can be a pleasant experience. The person 
who bases his security on overt independence is comforted by the 
satisfaction in flexing his muscles, admiring his own strength or beauty 
or wisdom, and reveling in his advantages over those whom he per- 


ceives as inferior. Adaptively self-confident individuals receive con- 
siderable admiration and social approval. 

In the maladaptive extreme, the narcissist seems driven to inflate 
himself compulsively at the expense of others. This brand of ab- 
normality leads to destructive activities. The severe narcissist cannot 
tolerate success or strength in others. He is driven to compete, to ex- 
hibit, to exploit. He is consistently rejecting and selfish. His com- 
pulsive and frantic attempts to boast lead to a most unrewarding circle 
of activities. As the narcissism becomes more flagrant, it fails to win 
respect; and this frustration leads to increased exhibitionistic maneu- 

The maxim of this form of maladjustment is: "How can 1 establish 
superiority over this person? How can I defeat him? How can I use 
him for my selfish enhancement?" 

The Effect of "22" Behavior 

Competitive, self-enhancing behavior pulls envy, distrust, inferior- 
ity feelings, and respectful admiration from others. In the language 
of the interpersonal system, BC provokes GHIJ. 

The adaptive person who uses this security operation in a sensitive 
manner wins the admiration and flattering envy of others. They look 
up to him and pay him the tribute of a grudging, envious approval. It 
may be helpful to contrast the interpersonal world created by the 
competitive person with the response which the executive, dominating 
person provokes. The latter is more conventional and responsible in 
his use of power. He tends to train others to obedience or loving 
respect. The competitive person strives to impress others that he has 
what they want. There is more disaffiliative motive in his approach 
and he generally receives therefore a passively hostile, negative sub- 

These generalizations are, of course, probability statements. The 
response of the "other one" is determined partly by his own inter- 
personal reflexes. Thus a rigidly docile, agreeable person may mani- 
fest the most friendly responses to a narcissist's approach. An inflex- 
ibly competitive person will react to another narcissist not with hum- 
ble defeated envy but with an increase in his own independent re- 
flexes. A fierce exhibitionistic competition between the two often re- 
sults. Thus the principle of reciprocal relations (which in this case 
reads BC pulls GHIJ) will be found to work in most cases but does not 
hold where inappropriate narcissism characterizes the subject or other 
rigid reflexes characterize the other. 


Clinical Manifestations of Co?npetitive Narcissism 

Narcissistic patients rarely come to a psychiatric clinic for diag- 
nosis or therapy. In one study of 537 routine admissions to a psychi- 
atric clinic only 6 per cent were diagnosed competitive or narcissistic 
at the level of symptomatic presentation. There are fewer narcissists 
in clinic samples than any other diagnostic type. 

The psychiatric clinic is thus not the natural habitat of the com- 
petitive, independent person. The reason for this finding seems clear. 
The emphasis on proud self-enhancement is quite incongruous with 
seeking psychiatric help. The very essence of this mode of adjustment 
is that "22's" ask help from no one, need no assistance, and are getting 
along quite well on their own steam. 

The "22" patients who do show up in the clinic generally come for 
one of three reasons: (1) psychosomatic symptoms, (2) current in- 
juries to their narcissism, (3) the desire to display their personalities 
or to talk about themselves. 

Patients who are referred to the clinic for ulcer or asthmatic 
symptoms often present independent, narcissistic fa9ades. The diag- 
nostic location for the average ulcer patient (at Level II-C) is in the 
BC (narcissistic-competitive) octant. Many of these patients tend to 
stress proud, hardboiled self-sufficiency. The same -is true of asthmatic 

The second group of competitive patients seen in the clinic are 
those whose self-regard has received a recent defeat. They often re- 
port the most colorful and fearful symptomology. They often list 
dozens of symptoms and may recount their eccentricities and life his- 
tories in great detail. The superficial impression of depression or de- 
pendence is deceptive. Psychological testing or perceptive interview- 
ing will reveal that the patients are not as anxious or depressed as they 
appear. What becomes evident is a narcissistic concern with their own 
reactions, their own sensitivities. The precipitating cause for their 
entrance to the clinic is usually a shift in their life situation, which 
causes frustration or a blow to their pride. The birth of a child may 
cause the narcissistic woman to become upset over the new demands 
of responsibility, nurturance, and the loss of attention. Narcissistic 
people in general react negatively to parenthood and intense conflicts 
may appear in this connection. One way in which this conflict can 
be handled is for the subject to incorporate the child into the circle 
of his or her own narcissism and thus share attention with the child. 

Occupational changes which lower public esteem or create de- 
pendency or require a tender approach may produce tensions in com- 


petitive men. Phallic, exhibitionistic men are often forced to retire to 
more sedentary, conventional occupations with accompanying pain 
and tension. This phenomenon was quite common after the last war 
when aviators, combat soldiers, etc., were faced with the loss of the 
grarificatiops of their positions. 

The histrionic character of this personality type often leads such 
individuals into activities which involve public display — modeling, 
acting, and other forms of social exhibition. Any shift in their life 
situation which involves the relinquishment of these rewards can 
make tension and symptoms and lead to psychiatric referral. 

Many competitive men are most comfortable when they are hold- 
ing independent positions — running their businesses, etc. Failures in 
those activities, which create a feeling of weakness or require them to 
take subordinate posts, can lead to increased anxiety and possibly 
physical symptoms. 

Another very common reason which brings the narcissistic persons 
into the psychiatric clinic is their intense interest in, concern for, and 
love of themselves. Many people perceive therapy as a unique oppor- 
tunity to talk about themselves, to spin theories about themselves, 
and to engage the interest and attention of a respected person (the 
therapist) in the subject that is dearest to the narcissist's heart. 

These three factors — physical symptoms, narcissistic injury, and 
self-fascination — seem to account for the motivation of those few com- 
petitive characters who come to the clinic. They lead to the paradoxi- 
cal situation of patients who are not really depressed or dependent 
applying for psychiatric help. 

Narcissists do not provide difficult problems for diagnosis if the 
intake worker focuses on the purposive meaning of their communica- 
tions and is not diverted by dramatic (but not deeply felt) symptomol- 
ogy. These patients are trying to impress the "other one." They may 
do this by muscle-flexing, boasting, seductive and colorful case 
histories, flirtatious maneuvers, or outright competition with the clini- 

There are certain psychometric signs characteristic of this personal- 
ity type. On the MA4PI, the anxiety and passivity scales (D and Ft) 
are low. The imperturbability scale (Ma) is high. Physical symptoms 
may push the Hs scale up. The conventionality-isolation scales are 
neither markedly high nor low. Thus F and Sc are not as high as in 
the case of the psychopath and schizoid. The conventionality scales 
K and Hy are not as pronounced as in the case of the psychosomatic 
and hysteric. 


Standard Psychiatric Definition of the Narcissistic Maladjustment 

In most of the preceding diagnostic chapters it has been possible to 
relate the interpersonal type of maladjustment to a standard psychi- 
atric category. Distrust defines the schizoid; docile dependency de- 
fines the phobic, etc. This relationship between interpersonal and 
psychiatric diagnosis does not hold in the case of the autocratic per- 
son, who has received relatively little attention from clinical theorists. 
These dominating, power-oriented persons do not tend to come for 
help, and have thus been neglected in the psychiatric literature. 

The same situation holds for the "22" personality. There is objec- 
tive evidence indicating that this personality type does not often come 
to the cHnic. There is, therefore, not a commonly agreed or Krae- 
pelinian-type term for categorizing these persons. 

This mode of maladjustment has, however, not been completely 
neglected by psychiatric writers. The psychoanalytic theory tends to 
focus not on the symptomatic factors (which are stressed by the pre- 
analytic psychiatrists) but stresses the multilevel aspects of character 
struct ure. This much more sophisticated approach considers character 
traits as means of warding off anxiety or instincts. The psychoanalysts 
have always recognized that self-love and independent narcissism form 
a common and eifective way of warding off or counteracting under- 
lying feelings of weakness. 

The term counterphobic is often employed to describe the exhi- 
bitionistic personality who compulsively attempts to demonstrate his 

Fromm has defined the exploitive character as one who attempts 
to better himself at the expense of others. 

Horney sees narcissism as one of the basic neurotic "trends." The 
narcissistic person, as defined by Horney, inflates himself and ag- 
grandizes himself at the expense of others. 

Jerome Frank and his colleagues have described three behavior pat- 
terns seen in psychotherapy groups which are very close to three of 
the interpersonal types presented in this book. The help-rejecting 
complainer and the doctor's assistant have been referred to in Chapter 
15 and Chapter 21, respectively. 

A third interpersonal type isolated by Rosenthal, Frank, and Nash 
(1, pp. 217-18) is called the self-righteous moralist. This mode of be- 
havior is quite similar to that being described in this chapter. These 
authors describe this type as follows: 

The most outstanding characteristic of the self-righteous moralist, as ex- 
emplified by these patients, is the need to be right or to show up the other 
fellow as wrong, particularly when some moral issue is involved which im- 
pinges on his own system of values, . . . 


In the very first group meeting, the self-righteous moralist tends to present 
himself as one who is calm, controlled, and self-contained, indicating his su- 
periority by a show of poise. He usually manages to become the focus of the 
discussion by his intensity, by dramatizing whatever he has to say, and by 
laboring his position indefinitely, refusing to concede any point, to admit any 
error, or to make any modification of his original formulation. . . . 

When symptoms, problems, and personal history are discussed, he talks of 
these in such a way as to enhance his own status: for example, he says that he 
has survived worse distress than others; that he has carried on in his duties 
despite his illness; that others are sicker than he is; and that others can profit 
from learning how he has handled his problems. . . . 

Schafer (2) has given more attention to the narcissistic personality 
than any other chnically oriented writer. As diagnostic cues he stresses 
"striking egocentricity," a tendency to avoid anxiety-arousing situa- 
tions (i.e., they do not like to exhibit behavior HI on the diagnostic 
circle). Schafer also mentions exhibitionism and overdemonstrative- 
ness, which he believes to be a cover-up of "basic coldness and dis- 
tance." In general it appears that the character disorder defined by 
Schafer is close to the narcissistic type of maladjustment described in 
this chapter. 

Research Findings Characteristic of the Narcissistic Personality 

In the preceding discussion of the narcissistic personality we have 
leaned upon and referred obliquely to research findings of the Kaiser 
Foundation project. Some of these results will now be summarized. 


Percentage of Competitive-Narcissistic 


(Level I-M) 

Found in Several Cultural Samples 


of Comp 


Institutional or Symptomatic Sample 



Psychiatric Clinic Admission 



College Undergraduates 



University Psychiatric Clinic 



Middle Class Obese Patients (Female) 



Overdy Neurotic Dermatids Patients 


Self-inflicted Dermatitis Patients 


Unanxious Dermatitis Patients 


Group Psychotherapy Pauents 


Individual Psychotherapy Patients 



Hypenensive Patients 


Ulcer Patients 


Medical Control Pauents 


University Counseling Center (Male) 


University Graduate Smdents (Male) 



Stockade Prisoners (Male) 



Hospitalized Psychotic Patients 
Officers in Military Service 







1. Patients who manifest competitive operations at Level I (MMPI) 
and in the rated interpersonal reflexes (sociometrics) do not have 
psychosomatic symptoms (except for the ulcer group). 

2. Ulcer patients are the psychosomatic group who stress com- 
petitive independence in their Level II self-descriptions. 

3. Competitive patients have MMPI profiles with low scores on 
depression (D) and obsessive rumination (Pt) and relatively higher 
scores on manic imperturbability (Ma). 

4. These patients do not tend to come to the psychiatric clinic. 
This diagnostic group is the least likely to accept a psychiatric referral. 

5. This personality type is found in other cultural samples more 
frequently than in the psychiatric clinic. As indicated in Table 28 
there are ten samples which contain more narcissists at Level I-M 
than the Kaiser Foundation clinic admission group. University gradu- 
ate students contain the most competitive personalities, followed by 
stockade prisoners and university psychiatric clinic patients.^ 

6. The Level II-C self-diagnoses of several samples are listed in 
Table 29. Comparison of Tables 28 and 29 is made difficult by the 
fact that the two samples which contained the greatest number of 
narcissists at Level I-M (graduate students and prisoners) were not 
included in the Level II-C study. At the level of conscious self- 
description, overtly neurotic dermatitis patients claim the most com- 
petitive self-confidence and the group-therapy patients (who em- 
phasize schizoid distrust) and self-inflicted dermatitis patients (who 
stress conventionality) claim the least narcissism. 

7. Narcissists are not especially motivated for psychotherapy. One 
sample of these patients came on the average for six therapeutic inter- 
views. This ties them for last place among diagnostic groups in terms 
of length of treatment. Female narcissists, incidentally, seem to stay 
in therapy longer than male narcissists. 

8. Competitive patients tend to be consciously disidentified with 
their parents. They are more identified than schizoids and psycho- 
paths but clearly less identified than the managerial, psychosomatics, 
hysterics, and phobics. 

9. The same findings hold for conscious marital identifications. 
10. Narcissistic patients tend to describe their parents as being 

relatively sadistic. 

* The fact that the military officer sample contains no narcissistic subjects would 
seem to be a contradiction to the previous statements which claimed that exhibitionis- 
tic characters are often located in military pursuits. The military officer group in- 
cluded here was tested under assessment circumstances which probably influenced 
their test-taking attitude and their resulting symptomatic scores. These officers were 
assessed in a nonclinical, quasi-military situation where there would be little pressure 
to stress narcissistic, unconventional feelings and some motivation to emphasize re- 
sponsible executive traits, which they did. 

% of Co?npetitive-Narctssisttc 



























Percentage of Competitive-Narcissistic Personalities (Level II-C) 
Found in Several Cultural Samples 

Instimtional or Sy?nptomatic Sample 
Psychiatric Clinic Admissions 
Hospitalized Psychotic Patients (Male) 
Group Psychotherapy Patients 
Individual Psychotherapy Patients 
Overtly Neurotic Dermatitis Patients 
Self-inflicted Dermautis Patients 
Unanxious Dermatitis Patients 
Medical Control Patients 
Ulcer Patients 
Hypertensive Patients 
Middle Class Obese Patients (Female) 

Total 781 

11. They describe their marital partners as agreeable, admiring 
people. This suggests that they tend to marry people whom they see 
as weak, docile, and who will pay them respectful tribute. 

12. On the Naboisek study of interpersonal misperception the com- 
petitive-exploitive group (along with the managerials) reveals marked 
misperceptions of weakness in others. They attribute too much passiv- 
ity and too much hostility to weak people. This suggests that an un- 
usually intense contemptuous superior attitude may exist in relation to 
weakness in others. 


1. Rosenthal, D., J. Frank, and E. Nash. The self-righteous moralist in early meet- 
ings of therapeutic groups. Psychiat., 1954. 17, No. 3, 215-23. 

2. Schafer, Roy. The clinical application of psychological tests: Diagnostic sum- 
maries and case studies. New York: International Universities Press, 1948. 


Adjustment Through Aggression: 
The Sadistic Personality 

The next sector of the diagnostic continuum is the area of critical hos- 
tile aggression. We shall consider in this chapter those human beings 
who manifest in their overt operations cold sternness, punitiveness, or 
sadism. This is the "33" personality. 

This way of life is traditionally one of the most fascinating and 
disturbing. We are dealing here with the fearful and destructive as- 
pects of human behavior. We shall attempt to understand why some 
individuals select negative, hostile expressions as their means of ad- 

Philosophers and psychologists have for centuries recognized that 
many human beings are compulsively committed to aggression. Many 
theories have been advanced to explain why some persons deUght in 
combat, feel comfortable only when engaged in a threatening attack, 
experience no qualms at punishing their fellows, and, indeed, feel 
weakened and threatened by the prospect of collaborative or tender or 
docile impulses. 

An important point must be introduced at this early stage of the 
discussion. We are referring in this chapter not just to actions of 
criminal aggression, destructive violence, or socially disapproved 
sadism. We include all those behaviors which inspire fear in others, 
which threaten others by physical, moral, or verbal means. 

Many antisocial individuals utilize this hostile mode of adjustment. 
But the great majority of punitive sadistic characters are to be found 
in the ranks of the socially approved. Those persons who consistently 
maintain a punishing attitude towards others, or a disciplinary atti- 
tude, or a sarcastic attitude, or a guilt-provoking attitude fall in this 
diagnostic category. Stern toughness is frequently admired and en- 
dorsed as a positive social adjustment. 



Those individuals who become repetitiously engaged in physical 
violence would, of course, be given the interpersonal diagnosis of ag- 
gressive personality. But a large percentage of the cases falling in this 
diagnostic category do not go around punching others — they com- 
municate their critical, hostUe messages in more subtle, but equally 
effective means. We think here, for example, of the stern unforgiving 
father, the bad-tempered wife, the moralistic guilt-provoking mother, 
the sharp-tongued mocking husband, the grim-faced punitive official, 
the truculent fiery-natured colleague, the disciplinarian. We include 
all those law-abiding, often pious and self-righteous, individuals who 
maintain a role of potential insult, derogation, or punishment. 

As we shall see in the subsequent sections, this mode of adjustment 
is far from being limited to the delinquent margins of society. It ap- 
pears with frightening regularity in the ruling groups of most societies 
— present and past — expressed in the philosophy of repressive legis- 
lation and bellicose foreign policies. 

In the next few pages we shall propose some speculations about the 
purpose, the effect, the survival advantages and disadvantages of ag- 
gressive security operations. 

The Purpose of "33" Behavior 

Those individuals who are overtly hostile and punitive have se- 
lected these behaviors because they sense them to be the most effective 
in minimizing anxiety. These interpersonal reflexes communicate a 
message of hardboiled toughness: "I am a dangerous, fearful person." 
The persons who rely on these operations for their emotional security 
are least anxious when they are flexing their muscles or expressing 
stern coldness. They are made most anxious in a situation which pulls 
for tender, agreeable, or docile feelings. 

These individuals have developed their involuntary interpersonal 
reflexes because they have learned consciously or unwittingly that this 
is, for them, the safest mode of adjustment. When they are acting 
tough or stern, they feel protected. When they act unaggressive, they 
feel unprotected and painfully uncomfortable. 

Sadistic, tough human beings apparently find security and pleasure 
in acting hardboiled; their self-respect seems to stem from the provoca- 
tion of fear in others. Hurtful, mocking, destroying, threatening ac- 
tions endow the actor with a fearful power. The threat of a temper 
outburst or a savage attack is a forceful weapon for coercing and 
managing others. Even the less violent aspects of this interpersonal 
operation — critical, disciplinary behavior — carry an authoritative so- 
cial weight. 


Hostile, critical conduct is generally viewed as negative and ethi- 
cally lamentable. Despite this moral censure this deportment is ac- 
companied by feehngs of righteousness. The most bitter delinquent, 
as well as the most punitive disciplinarian, often justifies his transac- 
tions by pious reasons. Sadists thus do not always feel the pain of guilt 
or the whip of social disapproval. In fact they often fit themselves into 
contexts where harshness and coercion are admired or accepted. The 
aggressive criminal gains respect in his own society. The martinet 
wins esteem within his own sphere of activities. The common genus 
of household sadist usually operates in reciprocal relationship to maso- 
chistic marital partners, who respond submissively. 

Another very important purpose of punitive or critical behavior is 
the provocation of guilt. The generic function of the hostile way of 
life is to destroy, to humiliate, to cow the "other one." This can be 
done violently. It can also be done indirectly. The cold, stern, dis- 
approving attitude has the aim of making the "other one" feel either 
inferior or unworthy. The sense of righteousness and austere punitive- 
ness is a most common and forceful attitude. Moral coercion is a most 
effective and self-satisfying form of sadism since it allows the release 
of destructive, hostile feelings along with the comforting support of 

The Effect of "3 3" Behavior 

We have seen that stern, hostile interpersonal reflexes serve several 
important purposes. They can reduce anxiety and the feeling of de- 
fenselessness. They express a feeling of armed protection, righteous 
irritation, and physical or moral superiority and force. 

This powerful social maneuver has quite a consistent effect on other 
people in general and certain rebellious or masochistic people in par- 
ticular. We shall consider first the general case. 

Sadistic-critical behavior pulls resentment, distrust, fear, and guilt 
from "others." In systematic language DE provokes FGH. 

The punitive, hostile role is a most effective interpersonal instru- 
ment. In the basic sense everyone fears destruction. Physical danger 
is, of course, the most crude and direct threat to any living organism. 
Social danger is, for the human being, a most fearful menace. This is 
expressed generically as disapproval or derogation. Almost everyone 
dreads and resents criticism and hostile laughter from others.^ 

Hostile coerciveness thus exerts a tremendous interpersonal lever- 
age. It gains a fearful respect or a resentful submission. Moral deroga- 

* An exception to this generalization is furnished by the overt masochistic charac- 
ter, cf. the discussion of the hostihty-provoking buffoon m Chapter 16. 


tion provokes guilt, and thus possesses a ruling force which can equal 
or surpass the threat of physical violence. 

In the moderate form, the critical role yields advantages to both 
the actor and the "other." No institution exists which does not depend 
to a certain extent upon social disapproval as a cementing and centrif- 
ugal agent. The stern, judicial, punitive person thus becomes a hu- 
man symbol of the rules and sanctions which exist either explicitly or 
implicitly. The critic or disciplinarian serves a healthy function in the 
economy of the group and gains respect and security for himself. 
Flexibility and adaptive moderation again become the criteria which 
differentiate the adjustive from the maladjustive. 

In many famihes one of the parents generally pre-empts this role, 
and thus gains the fearful respect and gives the reassurances of limits 
to the others. The well-adjusted aggressive-punitive person does not 
rely on these interpersonal reflexes rigidly. He can shift to other be- 
haviors when they are appropriate and when the critical functioning 
is not called for. 

In the extreme form, the sadistic role becomes the nucleus of com- 
plex neurotic phenomena. The maladjusted aggressive ; person is the 
one who manifests this operation inflexibly and to an intense degree. 
He operates as though anxiety is associated with the relaxing of tough- 
ness and this anxiety he cannot tolerate. 

Extreme or consistent sadism has a most electrifying effect on the 
"other one." Most people are made uncomfortable and ill at ease in 
the presence of an explosive or condemnatory or sarcastic person. 
They tend to fear him and to avoid him when possible. 

Most individuals can tolerate, and even appreciate, the function of 
an adjusted critic. They cannot tolerate potential or actual hostile 
coercion in others. This is to say that when extreme D behavior pulls 
adaptive withdrawal and bitter disaffiliation from "others," the inter- 
action terminates. 

There are two general occasions when this pattern does not hold: 
in the case of the reciprocal sado-masochistic relationship and in the 
crime-punishment partnership. There exists a large number of indi- 
viduals who are most comfortable (although not necessarily happy) 
when they are tied to a hostile partner. A most common variety of this 
is found in the masochists. An intense symbiotic relationship exists be- 
tween those who are least anxious when hurting or derogating and 
those who are least anxious when receiving these negative actions. In 
this case DE pulls intense maladaptive and rigid GH. 

Elaborate multilevel patterns exist in the sado-masochistic relation- 
ship. A maladaptive rigid sadistic fagade usually covers underlying 
feelings of fear and weakness. These are neutrahzed by the comfort- 


ing protection of hardboiled operations. The overt hostility in turn 
breeds guilt and a fear of retaliation which leads to an intensification 
of the original reflex. Similarly the overt masochist inevitably pos- 
sesses "preconscious" sadistic identifications. The masochist provokes 
hostility from the "other" which is generally followed by the provoca- 
tion of guilt in the aggressor. The anxiety associated with this hostile 
or righteous maneuver usually results in a resumption of the maso- 
chistic operations. 

The multilevel interactions of couples who are involved in sado- 
masochistic locks is one of the most interesting and complex human 
relationships. The delicate interaction between the two forms of 
sadism — physical and moral — are nicely illustrated in these not 
atypical cases. 

A second familiar symbiotic relationship exists in the intense recip- 
rocal partnerships between rebels and punitive authorities, between 
criminals and the agents of punishment. It is well known that irra- 
tionally unconventional and antisocial individuals pull hostility from 
others (cf. Chapter 15). The alienated schizophrenic, by means of his 
purposive eccentricity, provokes society to incarcerate him. The re- 
bellious student trains his teachers to discipline him. The professional 
radical eagerly searches his atmosphere for evidence of repressive 
cruelty (e.g., racial discrimination) and often succeeds, not in helping 
his cause, but in gaining the condemnation of others. Brilliant crea- 
tivity often reaches its peak in reaction to hostile, unsympathetic, re- 
strictive regulations. The other (punitive) side of this crime-punish- 
ment partnership works with equal purposiveness. Hostile punitive 
people seek out rebellious and distrustful others and integrate durable 
relationships with them. Policemen look for crime. The disapproving 
moral sadist looks for sinners. The bully feels most comfortable in 
receiving the resentful reactions of those he coerces. 

Remarkably intense and lasting relationships develop between anti- 
social rebels and the punitive figures whose anger they attempt to pro- 
voke. The recidivist criminal is least anxious when he is deaUng with 
the comforting consistency of prison custody. The punitive person 
is most comfortable when he has targets for his hostility. The severely 
maladjusted sadist thus gravitates towards bitter, guilty, and fearful 

Clinical Manifestations of the "33" Personality 

The sadistic type may be known by his symptoms. These pa- 
tients do not manifest the depressive characteristics of schizoids, ob- 
sessives, and phobics. They are not loaded with worries. They do 
not complain of physical symptoms. 


They come to the clinic usually under the pressure of unsatis- 
factory interpersonal relationships. Alarital problems are very com- 
mon. Discord and friction in their jobs frequently are mentioned. 
Often they are in trouble caused by their hostility. In these respects 
they appear much like the rebellious schizoid patients. Both diag- 
nostic groups emphasize negative, angry interpersonal reflexes. The 
schizoids are passively hostile, while the sadists are actively hostile. 
The schizoids are mad and sad; the sadists are mad and not sad. They 
are less concerned with their problems. As they describe their life 
events, a note of contempt and disgust with others often develops. 
It becomes clear that they are giving others in their life a bad time, 
that they are looking down contemptuously upon others. 

These patients often express unconventional ideas and admit to 
unconventional feelings. They may, in fact, make a point of avoid- 
ing conventional feelings and ideas, and when they do employ them 
they are often used to derogate others, A sadistic wife may, for ex- 
ample, admit to aggressiveness in herself and then criticize her hus- 
band for not being easygoing, A punitive man may admit his own 
sexual adventures with a certain hardboiled, sophisticated justification 
and wax indignant at the misconduct of others. 

In their demeanor during diagnostic interviews these patients gen- 
erally manifest aggressiveness coupled with some other interpersonal 
role. This is due to the fact that pure, unconflicted aggressive charac- 
ters rarely come to a psychiatric clinic. Invariably the aggressive pa- 
tient presents a conflicted fagade. His blunt, tough security opera- 
tions have led to trouble, or else he would not be visiting the clinic. 
A4any aggressive patients have intense covert feehngs of weakness or 
guilt. These may be apparent in the clinical interview. The brutal 
husband may express verbal guilt for beating his wife or children. 
The aggressive woman may verbalize pious conventional feelings in 
the effort to prove how contemptible her husband has become. The 
guilt in the first case and the bland conventionality in the latter will 
be seen to be superficial and verbal. These patients may sound guilty 
but they are not depressed. They may sound cooperative and agree- 
able but their contemptuous attitudes will be picked up by the Level 
I symptomatic tests (MMPI) or by the alert interviewer. These feel- 
ings often appear in the form of sarcastic or depreciatory references to 
psychiatry or psychotherapy. 

It has been pointed out that "33" characters come to the clinic 
complaining of interpersonal problems rather than anxiety symptoms. 
There are certain specific familial situations which are typical of this 
personality type. The blunt, active, righteous, angry wife of the 
delinquent husband is one such case. These patients often calmly 


describe a long history of marital turmoil in which a weak, immature 
spouse repeatedly offends the punitive wife with chronic alcoholism, 
gambling, unemployment, etc. The wife often supports the family 
and rules the spouse with a guilt-provoking disciplinary coldness to 
which the husband reacts with alternating guilt and rebeUion. These 
partnerships often are of long standing. The punitive member comes 
to the psychiatric clinic in the wake of a current episode of inter- 
rupted rebellion on the part of the spouse. The motive in coming may 
be to seek help in dealing with the husband, rather than help in 
changing her own behavior. The feeling of righteous indignation 
communicated becomes diagnostic. Inevitably testing reveals that 
masochistic trends underly these stern, punitive overt operations. 
"Preconscious" guilt and self -punishment picked up by fantasy tests 
often indicate that the patient is close to recognition of the underlying 
feelings of weakness, and these may provide the push which causes 
the patient to come to and stay in therapy. Another typical com- 
plaint of the aggressive character involves disgust or concern over 
symptoms in children. Delinquency, bed-wetting, and phobias often 
characterize the offspring of these patients. 

A third reason for coming to the clinic concerns authority prob- 
lems. The aggressive person often finds himself in a jam and comes 
for help under the pressure of disciplinary actions. These frictions 
are usually due to overharshness with subordinates, quarrels with 
equals, or insult to a superior. These patients do not manifest real 
guilt or unhappiness about these interpersonal conflicts, and they make 
it clear that the fault lies in the "other one." 

There are psychometric signs diagnostic of the stern or sadistic 
personality. Their MMPI profiles emphasize peaks on the hyper- 
mania, psychopathic deviate, and F scales. The Sc score is usually 
higher than Ft. The depression score is not pronounced. The Mf 
(femininity score) is usually low for male patients and varies for 
female patients depending on the amount of underlying masochism or 

Interpersonal Definition of the Psychopathic Maladjustment 

Evidence has been presented (cf. Chapter 12) that certain inter- 
personal maladjustive types were related to psychometric diagnostic 

Hostile, sadistic security operations are characteristic of the psycho- 
pathic personality. The essence of the psychopathic state is active 
aggression. These patients avoid anxiety and maintain security by 
avoiding dependent or tender feelings and by integrating critical, 
punitive relations with others. 



The classic generalization that psychopaths cannot love fits the 
looric of the circular diagnostic continuum since the DE octant which 
defines the psychopath is exactly opposite the affiliative sector of the 

Again it must be stressed that we are employing a definition of the 
kind and degree of abnormality which is based on personal, and not 
cultural, values. The cultural definition of the psychopathic maladjust- 
ment stresses the inability to conform to social norms. This is a poor 
definition because schizoid characters (as defined by the interpersonal 
system) seem to get into trouble as frequently as psychopaths. We 
have already stressed the point that many sadistic individuals are 
quite acceptant of punitive and repressive ethical values. The psycho- 
pathic personality in the interpersonal system is defined by the afore- 
mentioned typical security operations and not by delinquency. As 
a matter of fact, sadistic people are perhaps more often unusually 
identified with law and moral codes which they ruthlessly employ to 
humiliate others. The more a person goes out of his way to claim an 
ethical superiority and to attribute immorality to others, the greater 
the probability that he manifests psychopathic, morally sadistic 

The distrustful schizoid patient is acutely aware of moral hypocrisy 
in others. Some psychopaths often show a radar-like sensitivity to 
rebelliousness and guilt or weakness in the "other one." 

Research Findings Characteristic of the Sadistic Personality 

Some of the current research findings which concern the aggressive 
personality can now be considered. 

L Patients who manifest stern aggressiveness in their overt opera- 
tions do not have psychosomatic symptoms. 

2. Psychosomatic patients do not utilize these interpersonal opera- 
tions at Levels I and IL 

3. Aggression at Levels I and II is related to high MMPI scores on 
nonconformity (F), schizoid distrust (Sc), and disidentification with 
affiliative values (Pd). 

4. If sadistic patients enter psychotherapy, they tend to remain 
in treatment for long periods. They stay in therapy as long as any 
other diagnostic type. Severely conflicted psychopaths (i.e., large 
discrepancies between Level I and II) are, however, poorly moti- 
vated for treatment, remaining on the average for only two sessions. 
This indicates that the psychopath like the phobic presents a tricky 
prognostic gamble. They either avoid therapy entirely or they enter 
and remain for extended periods. Multilevel conflicts can lead both 
of these groups to avoid intensive treatment. Schizoid and obsessive 


patients on the contrary are more likely to stick in treatment regard- 
less of the degree and kind of multilevel conflict. 

5. The psychopaths, like the schizoids, are the most disidentified 
(consciously) with their parents. 

6. They tend as a group to be consciously disidentified with their 
marital partners. 

7. They (along with the schizoid group) tend to misperceive 
the interpersonal behavior of others. They inaccurately attribute too 
much hostility to others. 

8. Patients who manifest aggressiveness in their Level I-M sympto- 
matic behavior tend to appear in certain cultural samples much more 
frequently than others. The percentage of sadistic persons in various 
samples is presented in Table 30. 


Percentage of Aggressive-Sadistic Personalities (Level I-M) 

Found in Several 

Cultural Samples 


of Aggressive-Sadistic 

Institutional or Symptomatic Sample 



Psychiatric Clinic Admission 



College Undergraduates 



University Psychiatric Clinic 



Middle Class Obese Patients (Female) 



Overtly Neurotic Dermatitis Patients 



Self-inflicted Dermatitis Patients 



Unanxious Dermatitis Patients 



Group Psychotherapy Patients 



Individual Psychotherapy Patients 



Hypertensive Patients 



Ulcer Patients 



Medical Control Patients 


University Counseling Center 



University Graduate Students (Male) 


Stockade Prisoners (Male) 



Hospitalized Psychotic Patients 



Officers in Military Service 




As might be expected, the stockade prison group contains the 
largest percentage of psychopathic individuals. The psychiatric 
clinic samples include about the number expected by chance. All the 
other samples have a negligible number of this personality type. 

It is of interest to note the discrepancy between psychopathic per- 
sonalities reporting to the clinic and those going into group or indi- 
vidual psychotherapy. Only half the expected percentage of sadistic 
individuals go into treatment. This indicates that either those pa- 
tients tend to avoid going into therapy or the clinic refrains from re- 


ferring them to therapy. They come to the clinic in the expected 
frequency but their reasons for coming (which often involves blam- 
ing others) do not lead them to go into therapy. Once they do enter 
treatment, they tend to stay a relatively long time. This means that 
sadistic patients are poorly motivated but have long prognosis for 

9. The percentage of sadistic personalities at Level II-C is pre- 
sented in Table 3L Ulcer patients claim the most aggressiveness. 
Normal controls and psychiatric clinic admissions attribute more 
sadism to themselves than expected by chance. The hypertensive and 
obese samples (who stress hypernormal strength) have considerably 
fewer self-diagnoses in the aggressive-sadistic octant. 


Percentage of Aggressive-Sadistic Personalities (Level II-C) 

Found in Several Cultural Samples 


oj Aggressive-Sadistic 

Institutional or Symptomatic Sample 



Psychiatric Clinic Admissions 



Hospitalized Psychotic Patients (Male) 



Group Psychotherapy Patients 



Individual Psychotherapy Patients 



Overtly Neurotic Dermatitis Patients 



Self-inflicted Dermatitis Patients 



Unanxious Dermatitis Patients 



Medical Control Patients 



Ulcer Patients 



Hypertensive Patients 



Middle Class Obese Patients (Female) 





10. Sadistic patients describe their parents as weaker and more 
neurotic than any other diagnostic group. They clearly look down 
on their parents — seeing them as beaten, impotent, timid, unsuc- 
cessful, and unloving people. 

1 1 . Sadistic patients describe their spouses as rebellious and resent- 
ful people. This indicates that their marital relations are loaded with 
punitive affect. They are stern and disapproving in relation to resent- 
ful, passively resistant spouses. 


Some Applications of the Interpersonal System 


The aim of the Kaiser Foundation research has been to develop a 
system of personality which is functionally useful in the psychiatric 
clinic. The norms have been based on clinic samples. The empirical 
investigations have attempted to build up probability statistics which 
allow us to predict what the patient will do in the clinic setting. 

The system has, for the first six years of research, been deliberately 
restricted for the most part to that narrow range of interpersonal be- 
havior which is relevant at the time of intake evaluation and plan- 
ning for psychotherapy. The major appUcation of the system is to 
problems of diagnosis and prognosis faced in the psychiatric clinic. 

While the main focus has been on cUnical diagnosis, there have 
been some side explorations to determine the efficacy of the inter- 
personal system in predicting behavior outside of the psychiatric 

In the next four chapters we shall report on some applications of 
the interpersonal system in these nonclinical situations. 

Chapter 23 will report on the use of these diagnostic methods in 
a psychiatric hospital. The interpersonal pressures faced by the staff 
of an inpatient service are clearly different from the outpatient fa- 
cility. Our experience in using the interpersonal methodology in 
this situation is very limited. The results of our pilot study are strik- 
ingly different from the outpatient studies and are partially con- 
firmed, moreover, by our measurements on outpatients who have had 
psychotic breaks and required hospitalization. Some suggestions and 
impHcations concerning the apphcation of interpersonal diagnosis in 
the psychiatric hospital will be tentatively advanced. 

Chapter 24 takes us to a different environmental setting — the phy- 
sician's office. We shall study the application of the system to psycho- 
somatic problems faced by the internist and the dermatologist. Four 
samples of patients manifesting symptoms which are sometimes be- 
lieved to be psychosomatic have been studied by the Kaiser Founda- 
tion research. The diagnostic system suggests that there are some 
personality correlates of psychosomatic conditions. The results of 
these studies and their clinical implications for the physician will be 




Chapter 25 describes some interpersonal dynamic factors observed 
in administrative, discussion, and management groups. The inter- 
personal measurement methods are easily converted into sociometric 
instruments. Patterns of reciprocal interpersonal relations and mis- 
perceptions of self and others are measured by very straightforward 
techniques. They provide a direct method for diagnosing an industrial 
management group and outlining the network of dynamic activities 
which occur in the group situation. 

In Chapter 26 this survey of the application of the diagnostic sys- 
tem concludes with a consideration of interpersonal dynamics as they 
occur in group therapy. Methods for predicting and measuring group 
resistance and group personality will be described. This chapter also 
discusses the complex issue of multilevel interaction patterns as they 
evolve in group psychotherapy. 


Interpersonal Diagnosis of 
Hospitalized Psychotics 

The emotional atmosphere in any psychiatric hospital is inevitably 
different from that of the outpatient clinic. Implicit interpersonal 
forces are at play which affect the patient's behavior. One task of a 
diagnostic system employed in the hospital setting is to measure the 
patient's reactions to the social pressures of the hospital environment. 
The standard questions to be answered by a multilevel measurement 
apparatus are: What overt social role is the patient attempting to 
maintain (i.e., what are his interpersonal security operations)? How 
does he diagnose himself? What are his underlying feelings? 

The interpersonal diagnostic system has been used in some limited, 
exploratory studies in a hospital setting. Some of the results will be 
reported in this chapter. These findings are preliminary and sug- 
gestive. They do contain several implications about the use of per- 
sonality tests in the hospital and about the nature of the psychotic state. 

Factors Unique to the Psychiatric Hospital 

In assessing the interpersonal behavior of hospitalized patients it is 
clear that certain factors peculiar to institutional commitment are 
involved. The patients are certainly involved in a different relation- 
ship with the therapist than are patients in the outpatient clinic. The 
latter come mainly on their own volition. The clinic does not play 
such a vital 24-hour-a-day role in their lives. The outpatient is much 
more free to sever his relationship with the clinic. 

The hospitalized patient is inextricably caught in a web of inter- 
personal assumptions which affect his behavior and his conception 
of self. He is legally committed; he has been rejected by society, and 
often by his family. He is not as free to govern his actions. He is de- 
pendent on the institution for sustenance, both material and emotional. 




The interpretation of test results must take into account the emo- 
tional context of the hospital scene. We can never be sure how much 
the patient's behavior is directed towards the therapist-as-therapist and 
how much it is determined by his attitudes and interpersonal purposes 
towards the custodial institution and the rejecting outer world to 
which the hospital is related. 

There is another factor which limits the application of the inter- 
personal system to hospital diagnosis. The system was developed to 
meet the needs of patients in a clinic. It is geared to patients who are 
of average intelligence and who are able to manage their affairs by 
themselves. Individuals who are severely psychotic (i.e., out of touch 
with reality, which usually means wildly rebellious against conven- 
tional standards) may not be able to meet the intellectual demands of 
our check lists and questionnaires. 

For this reason we are very cautious in recommending the inter- 
personal system for general use in the psychiatric hospital. For many 
patients it seems to work with adequate success — that is, it success- 
fully answers the question: what are the patient's interpersonal ac- 
tions, beUefs, and underlying feelings? 

The Psychotic Samples 

Three samples of psychotic patients were studied by the inter- 
personal diagnostic system. 

The State Hospital Sample ^ comprises 100 patients tested at Level 
I-M and 6 patients who were referred for group psychotherapy and 
were administered the multilevel interpersonal test battery before be- 
ginning treatment. The 100 patients were a random sample of patients 
who received the MMPI during diagnostic work-ups. 

The criteria for selecting the six other patients were as follows: 
Seven patients were assigned to the therapy group. Six of them took 
the tests and received interpersonal diagnoses. The seventh patient 
was too disturbed to respond to the testing situation. All group mem- 
bers came from the same unit, a convalescent cottage. Five had re- 
ceived diagnoses of schizophrenia; one was diagnosed as a depressive 
psychotic. None were currently receiving any somatic treatment. 
The cottage is a semi-open ward and does not contain acutely dis- 
turbed patients. While the rate of discharge of patients in this unit is 
relatively high, patients selected for this therapy group were not ex- 
pected to be leaving the hospital within the next four months. 

* The sample of six hospitalized patients was collected by Richard V. Wolton of 
the Stockton State Hospital, Stockton, California. Gratitude is expressed to Mr. Wol- 
ton for his cooperation in administering the tests and for writing the clinical sum- 
maries included in this chapter. 


A second sample of hospitalized patients has been studied by the 
interpersonal diagnostic system. These comprise patients who were 
evaluated in an outpatient clinic, diagnosed either as psychotic or 
anxiety-panic types and then hospitalized. This group of patients is 
called the Clinic Psychotic Sample. 

There are 22 subjects in this "panic-psychotic" sample, 4 men and 
18 women. Only 13 of these patients completed the tests at all three 
levels so that the N's vary from level to level. 

The Private Hospital Sample includes 20 patients who were in 
psychotherapy at the Pinel Foundation Hospital, Seattle, Washing- 
ton.^ There are obvious cultural and clinical factors which might dif- 
ferentiate private hospital patients from those seen in a state hospital 
or a health-plan chnic, but all three samples share the common ex- 
perience of having been institutionalized because of this emotional 

Level I Behavior in the Three Psychotic Samples 

Level I-M scores are available for the state hospital and clinic 
psychotic sample. Level I-S ratings of each patient by professional ob- 
servers (pooled ratings of doctors, nurses, and therapist) are available 
for the private hospital sample. Table 32 presents the number of pa- 
tients falling in each diagnostic category at Level L 

Level I Diagnoses Assigned to 148 Patients in the Three Psychotic Samples 



Level I Diagnosis 
12 3 4 5 6 7 8 






State Hospital Sample . 

Clinic Sample 

Private Hospital Sample * 

Total Sample 




22 17 15 6 6 10 6 24 
2 2 3 4 3 3 14 
12 3 5 7 2 

25 21 21 10 14 20 9 28 





* The figures presented for the private hospital sample are based on Level I-S 
(raters' sociometric checks). The other two samples were studied at Level I-M— the 
symptomatic level. 

No clear-cut trend is evident in these results. The findings sug- 
gest that hospitalized patients present a mixed picture at the sympto- 
matic or overt interpersonal level. When the diagnostic continuum 

^ Gratitude is expressed to the administration of the Pinel Foundation Hospital and 
to Dr. Arthur Kobler of the Pinel Staff for permission to use the results obtained in 
their diagnostic studies. 



is summarized in terms of positive conventional types (1678) as com- 
pared with hostile alienated types (2345), there are slightly more 
psychotic patients in the former category. In light of the nature of 
this sample this becomes a most interesting result. A4ore than half of 
these patients who have been rejected by society and institutionalized 
for emotional disturbance present themselves as responsible, hyper- 
normal, or conforming people. The implication is that many psy- 
chotics strive to maintain a Level I fagade of conventionality and con- 

The findings listed in Table 32 tend to duplicate the census of pa- 
tients in the Kaiser Foundation outpatient cUnic. These results sug- 
gest that the interpersonal pressure of the symptoms of psychotic pa- 
tients does not differ from that manifested by the average outpatient 
visitor, and (as we shall see in the next chapter) the Level I fagade 
of psychotics is significantly more hypernormal than that of out- 
patients who go into psychotherapy. 

Level II-C Behavior in Three Psychotic Samples 

The Level II-C self-diagnoses of the hospitalized patients which 
are presented in Table 33 present an even more interesting pattern. 

Level II-C Diagnoses of 46 Patients in the Three Psychotic Samples 



Level II-C Diagnosis 
12 3 4 5 6 7 8 





State Hospital Sample . . 

Clinic Sample 

Private Hospital Sample . 

Total Sample 




2 3 10 
2 13 4 2 2 3 3 
4 12 116 14 

8 5 5 5 3 9 4 7 







There is a definite tendency for institutionalized patients to see 
themselves as sweet, hypernormal, executive individuals! They tend 
as a group to deny hostile or weak traits. This strongly suggests that 
many psychotics cling to a conscious self-perception of strength, re- 
sponsibility, and conventionality. They are significantly less ahenated 
and guilty (at Level II-C) than outpatients in psychotherapy. 

Level lll-T Behavior in Three Psychotic Samples 

The TAT "hero" indices were calculated for 38 patients in these 
studies and then plotted on the diagnostic grid. Table 34 presents 
the Level III-T diagnoses for the three samples. 



Level III-T Diagnoses of 38 Patients in the Three Psychotic Samples 



Level III-T Diagnosis 
12 3 4 5 6 7 8 





State Hospital Sample . . 




2 12 1 
3 2 3 10 2 1 
3 3 5 3 2 2 2 

6 3 9 6 4 4 2 4 





Private Hospital Sample . 
Total Sample 



At the "preconscious" level these patients manifest interpersonal 
themes which are somewhat different from their overt and conscious 
presentations. Sadistic themes are the most common; power and dis- 
trust are the two next most frequent. Whereas the psychotics were 
significantly more bland and hypernormal than psychotherapy out- 
patients in conscious self-description, in fantasy they are significantly 
more hostile. 

Multilevel Personality Patterns of the Hospitalized Samples 

It is now possible to weave together the results from the three levels 
of personality and to make multilevel summary statements. 

At the symptomatic level (Level I-M) the psychotics equal 
psychiatric clinic outpatients in the presentation of symptoms. At 
the level of conscious self-diagnosis a larger percentage of psychotics 
claim strength and conventional normality than do outpatients. At the 
level of "preconscious" fantasy, however, the psychotics are more 
bitter and hostile than the outpatients. 

These data have suggested the following hypothesis. Many psy- 
chotics show highly conflicted personality patterns. Their overt se- 
curity operations emphasize strength and normality, while their under- 
lying feelings involve sadism and bitterness. Many psychotics cling 
desperately to a conscious fa9ade of conventionality and self-confi- 
dence in the teeth of their underlying feelings of rage and frustration. 

The presenting operations of many neurotics, on the contrary, 
stress overt passivity and bitterness, while their underlying feelings 
involve stronger and more affiliative feelings. 

The over-all impression obtained from these studies is that the 
process of hospitalization involves different factors. In looking over 
the multilevel diagnostic codes for the individual patients, it is apparent 
that many patients are institutionalized because they are immobilized 



by anxiety, distrust, guilt, and helplessness at all three levels. These 
would be clinically labeled anxiety-panic states, schizoid conditions, 
psychotic conditions, or psychotic depressions. Another larger group 
of hospitalized patients try to maintain a fa9ade of strength and to 
cover up intense underlying feelings of rage and bitterness. The for- 
mer group probably includes suicidal risks, w^ithdraw^n and apathetic 
operations. The latter group are usually called paranoid. Different 
therapeutic implications exist for these two broad groups of insti- 
tutionalized cases. 

Implications of the Multilevel Patterns of Psychotic Patients 

A multilevel system of personality throws into clear relief the con- 
flicts which exist in a patient's character structure. We have just re- 
viewed two sets of evidence suggesting that many psychotics and pre- 
psychotics present a two-layer facade of strength and normality 
covering intense feelings of rage and despair. There are several im- 

The first concerns the therapeutic handling of psychotics. Most 
hospitalized patients are institutionalized because they have exhibited 
unconventional behavior which frightens or alienates others. They 
are seen by others as crazy, psychotic, disturbed. 

Many of these patients, however, see themselves quite differently 
— they strive to present themselves as confident and responsible. It 
seems clear that anyone who attempts to establish constructive com- 
munication with this kind of psychotic must pay respect to the fa9ade 
of normality. The overt operations must always be understood and 
classified before therapy can deal with underlying feelings. Patients 
(in the cUnic or in the hospital) who present a fagade of strength 
tend to be poorly motivated for psychotherapy — since treatment 
threatens their security operations. Extended and painstaking pre- 
liminary procedures (educational talks, discussions in which the 
therapist stays on the side of the ego) may be necessary to prepare 
such patients for conventional psychotherapy. If these are short- 
circuited and an attempt is made to plunge the patient into therapy a 
disastrous communication situation develops — in which the therapist 
acts as though the patient needs treatment and the patient thinks and 
acts on the premise that he does not. 

This multilevel psychotic profile has implications for the use of 
psychological tests. Many psychologists have used Level I and II 
instruments in testing institutionahzed psychotics and have been dis- 
appointed in finding that the patients appear normal. The allegation 
has been made that the MMPI is invahd because it often reveals psy- 


chotics as having normal profiles. Much confusion and damage has 
been caused by researchers who have administered tests to patients 
with a unilevel point of view. 

If test responses are viewed as interpersonal communications be- 
tween the patient and the psychologist and if a multilevel approach is 
maintained, then the issue of validity becomes clarified. If a psychotic 
produces a low MMPI profile and describes himself on questionnaires 
as nonsymptomatic, this does not invalidate the tests. On the con- 
trary, these results provide most useful information. They tell the 
tester that the patient is attempting to maintain a fa9ade of normality, 
that he wants to be seen as healthy and nonneurotic. The conception 
of levels enters here. The sophisticated diagnostician will proceed to 
administer tests which tap other levels. He will compare the overt 
and conscious "normal" operations with Level I reports from observ- 
ers that the patient acts hostile, or with Level III materials indicating 
that intense private feelings of distrust exist. The bland fagade is 
seen to cover paranoid hostility or psychotic despair. 

The same unilevel error has been made by research psychologists 
who have attempted to demonstrate that the TAT fantasy story test 
is invalid because it does not differentiate between neurotics and 
psychotics. It is very possible that a sample of psychotics will show 
no more hostility and weakness in their TAT stories than neurotics. 
If these investigators had gone on to collect measures at the overt 
levels, they would have been surprised to discover that more psy- 
chotics covered their fantasy behavior with a fa9ade of normality 
than did neurotics. The comparison between any two clinical groups 
cannot be made at a single level of personality without confusion or 
incomplete results. 

The multilevel pattern which we have found to characterize many 
psychotics has implications for evaluating outpatients for psycho- 
therapy. Ad any patients are seen for intake evaluation in the Kaiser 
Foundation clinic who manifest the multilevel pattern of overt con- 
ventionality with underlying sadism or distrust. There are dozens of 
such multilevel diagnoses, e.g., 773, 884, 173, etc. This is a prepsy- 
chotic pattern. These patients, many of whom are presenting psycho- 
somatic symptoms, are never assigned routinely to therapy or analysis. 
In many cases they are given the opportunity to "repress out" of 
therapy or are seen in supportive counseling. If a patient with a pre- 
psychotic multilevel diagnosis is assigned to treatment, the therapist 
is warned to watch for signs of anxiety. The question is posed: can 
this patient tolerate conscious awareness of his underlying feelings? 
Reactions to the earliest interpretations should be observed to see if 
psychotic trends or "flights into health" are developing. 


Case Illustrations of Six Psychotic Patients 

In order to illustrate the use of the interpersonal diagnostic sys- 
tem in the psychiatric hospital, we shall now present the test results 
and clinical histories of the six patients from the state hospital sample. 

In each illustration we shall consider first a brief case history and a 
description of the patient's behavior in the group. The interpersonal 
diagnostic report will follow. It will be possible to compare the test 
results with the way the patient behaved both outside the hospital 
and in his therapy group. 


(1) Clinical Data. This 42-year-old male patient was born in a 
rural Midwestern state. His history is one of marginal social adjust- 
ment, marked by many arrests for drunkenness, molesting children, 
vagrancy, and nomadism. He was committed to this hospital shortly 
after his arrival in the community by freight train. At the time of his 
hospitalization the patient heard voices directed by the church, felt 
that the cabin in which he resided was wired with microphones to 
find out his thoughts, felt that medicine was poisoned. He impressed 
the examining physician as "friendly but shy." The diagnosis ad- 
vanced was schizophrenic reaction, paranoid type, in a constitution- 
ally inadequate individual associated with chronic alcoholism. Electro- 
shock therapy was initiated shortly after the patient's admission to the 
hospital, and during this course of treatment the patient was involved 
in numerous "special incident" reports and as a consequence was fre- 
quently placed in seclusion and restraint. After twenty-four electro- 
shock treatments, the patient was transferred to an open ward where 
he had ground privileges. He was a member of the therapy group 
for four months. Four months after the group terminated the pa- 
tient's paranoid symptoms returned, and he was transferred to a 
closed ward where a second series of electro-shock treatments was be- 
gun. After four such treatments there was a moderate improvement 
and electro-shock therapy was discontinued. At last report the patient 
was doing well and has had his ground privileges restored. 

(2) Group Behavior. The patient missed a few of the early meet- 
ings of the group because of his inability to remember days of the 
week. He knew, for example, that meetings were held each Wednes- 
day at 9:00 a.m., but did not know when it was Wednesday. The 

^^ Certain changes in peripheral details have been made in these case histories in 
order to insure anonymity. An attempt to preserve the essential quahty of the case 
history has been made. The descriptive titles for each patient were supplied by the 
clinician who conducted the psychotherapy group. 


patient's speech, especially in early meetings, was circumstantial and 
rambling; his manner was vague, nebulous, and cloudy. He seemed 
uncertain of his identity and seemed to have very few resources in the 
way of stable and enduring personaUty characteristics. His adjust- 
ment appeared to be on a day to day and even on a minute to minute 

At about the seventh meeting of the group the patient began to 
change. He took a more active part in group discussions, no longer 
missed meetings, and displayed obvious interest in the responses and 
reactions of the others in the group. Although profoundly amnesic 
to many episodes in his past, the patient seemed to be trying to put 
the pieces together and re-establish and redefine his personality out- 
lines. In early sessions the patient's verbal responses were character- 
istically platitudinous, while in later sessions his comments seemed 
honest, direct, and at times, poignant. When the group terminated, 
it was noted that while the patient was considerably improved, no 
change in his hospital status was recommended. 

(3) Interpersonal Diagnostic Report. This patient at Level I-M 
presents as mildly depressed — somewhat despondent. He does not 
stress his symptoms, does not attempt to make an extremely sick or 
neurotic impression. He is diagnosed as a moderate phobic or de- 
pendent personality at this level. This seems to fit the clinician's im- 
pression of his "friendly but shy" approach. 

In his conscious self-description he presents a mixed picture. He 
denies hostility. He admits to some passivity but also claims inde- 
pendence and strength. He is very close to his ego ideal — indicating 
that he is self-satisfied, pleased with his personality, and not motivated 
for psychotherapy. 

His "preconscious" hero themes stress inordinate passivity and feel- 
ings of weakness. 

The multilevel pattern thus reveals a strong, independent fa9ade 
with some feehngs of depression — and underlying feelings of helpless- 
ness. His basic feelings of despair and weakness are expressed in- 
directly in his symptoms but are not consciously recognized. These 
underlying feelings apparently did reach expression in the poignancy 
noted in later sessions of the group. 

(4) Clinical Implications, (a) Motivation. This patient would 
not be considered as well-motivated because he is not under great 
symptomatic pressure (Level I-M = moderate 6) and is self-satisfied 
(Level II-C = i). 

(b) Prognosis. Prognosis is complicated by the underlying feel- 
ings of helplessness and dependence (Level III-T = 6). Male pa- 


rients whose "preconscious" themes locate in this octant are more 
difficult to treat because therapy will lead to the expression of passive 
(and usually feminine) emotions. 

(c) Predicted interpersonal behavior. The Level I-M and Level 
II-C measures tend to predict accurately to the platitudinous approach. 


(1) Clinical Data. At the time of his admission to the hospital 
this patient was a 42-year-old white married male who spoke in an 
irrelevant, illogical manner much of the time. He showed many re- 
ligious delusions and was depressed and agitated. The patient had 
been married for the past fourteen years, and throughout this period 
showed an abnormally strong attachment to his father. He has fre- 
quently expressed a desire to help his father at the expense of his wife 
and two children, who were often in dire financial straits. He would 
become violently angry if thwarted in his desire to aid his father, 
and on one occasion knocked out several of his wife's teeth when she 
expostulated with him. He frequently expressed ideas that he should 
make his living as a writer, although he has had nothing published. 
In the two months immediately prior to his commitment the patient 
became much more disturbed and confused. He would preach con- 
stantly and incoherently, stating that God had directly communi- 
cated with him. 

The patient was given electro-shock therapy and made an im- 
mediate and favorable response to it. After six treatments the patient 
was much improved. He became a member of the therapy group, and 
after the group terminated the patient was given an indefinite leave 
of absence to his family. On subsequent examinations at the hospital 
the patient was described as "sullen and aggressive" by the examining 
physician, but there had been no relapse of sufficient degree to war- 
rant hospitalization. 

(2) Group Behavior. The patient's behavior in the group was 
characterized chiefly by his sober, earnest manner. He rarely smiled, 
and was by far the most reflective member of the group. From the 
beginning, the patient's contributions to the group discussion M^ere 
relevant, pertinent, and coherent, with no evidence of the psychotic 
manifestations contained in the commitment report. The patient was 
quite self-punitive in presenting his problems to the others in the 
group. He described himself as a failure as a father, as a husband, as a 
person, as a writer. He had much to say about how he always felt 
he should be a writer; how he always admired the use of language 


and, especially, "big words." On one occasion he brought a collec- 
tion of his writings to the group. These were all written when the 
patient was in late adolescence, and the papers on which they were 
typed were crinkled and abused by age. In substance, they contained 
a very ponderous philosophy in poetic forms, reminiscent of the duller 
works of the Victorian period. 

The patient revealed a concern for the problems of the others in 
the group. He was supportive to an indiscriminate degree, and seemed 
to be asking for support when he chastised himself as a failure before 
the others. Such support was not reciprocated by the other patients, 

(3) Interpersonal Diagnostic Report. This patient presents at 
Level I-M a mildly depressed, essentially normal picture. (All MMPI 
scales are below 70.) He is definitely not attempting to impress 
others as a sick, nervous person. He is diagnosed at this level as a do- 
cile personality. 

In his conscious self-descriptions he stresses strength and hyper- 
normal responsibility. He is quite close to his ego ideal — indicating 
self-satisfaction and no awareness of any need to change his personal- 
ity- . ^ 

The top two levels thus indicate a normal, conventional self-conn- 
dent facade. 

At the level of fantasy a different picture develops. Intense feel- 
ings of bitter distrust and aggressive power are expressed. 

The three-level pattern involves two layers of normality (he is a 
strong man, mildly depressed) covering intense sadistic feelings. 

When this multilevel pattern is compared with that of the preced- 
ing patient, we observe that they are quite similar in their facades 
(both 61) but very different at Level III-T. The first patient ex- 
pressed helpless fantasies, which we related to his poignancy. The 
second patient manifests bitter, angry feelings, which are reflected in 
the sadistic violence reported in his chnical history and, perhaps, in 
the fact that the group responded negatively to him. 

Once again we see a common psychotic pattern of a frail fa9ade 
of normality conflicting with underlying pathology. 

(4) Clinical Implications, (a) Motivation. This patient would be 
considered unmotivated because of the symptom-free, self-satisfied 

(b) Prognosis. The prognosis is complicated because of the in- 
tense "preconscious" feelings of hostility. The conflict between a 
bland fagade and underlying bitterness (61^) is always a potentially 
explosive one and difficult to treat by psychotherapy. 


(c) Predicted interpersonal behavior. The three-level diagnostic 
code predicts the earnestness, the pedantry, and the sullen outbursts. 
It does not pick up the masochistic self-derogation — unless this is 
interpreted as complaining hostility. 


( 1 ) Clinical Data. The patient was committed at the age of thirty- 
three by his wife and his mother following a suicide attempt (sleeping 
pills). The patient had asked his wife to join him in a suicide pact 
and had expressed feelings of hopelessness and profound despair for 
the few weeks immediately preceding his hospitalization. The cUni- 
cal decision was that his primary diagnosis was a reactive depression, 
but of such a severe nature that it approached psychotic manifesta- 
tions. The patient adjusted quite well to the hospital milieu from the 
beginning. He was assigned to an open ward and worked days in 
the same type of work in which he was employed before hospitali- 
zation. No somatic therapy was deemed necessary or advisable. The 
patient entered the therapy group and remained in it for six sessions. 
Then he left the hospital without permission. The patient's elope- 
ment was sudden and unexpected, since he had always been reluctant 
to discuss being discharged from the hospital. The patient frequently 
stated that he was rather afraid to leave the hospital to return to a 
world in which all sorts of terrible things could happen to him. Since 
the patient's unauthorized departure, no word has been received by 
the hospital regarding him. 

(2) Group Behavior. Until the time of his abrupt departure from 
the hospital the patient was an active participant in the therapy group. 
A quiet man by inclination, he was very attentive to topics of discus- 
sion, listening with alertness and active interest. He encouraged other 
patients to discuss their problems in the group setting, but found it dif- 
ficult to lead the way by using his own case as an example, although 
this was his expressed intention. The patient was probably in a better 
state of mental health than any other group member, and his partici- 
pation directly reflected this. When another patient would express 
delusional material, he would try to steer the conversation into more 
comfortable channels. He was well liked by the other members of 
the group, who would refer to him as a "nice guy." The patient was 
discussed more freely by the others following his elopement than he 
was during the period in which he was an active group member. 
Resentment against his blandness and manner of departure emerged, 
was discussed, and was related to the patient's attempts to control 
group discussions and keep them on a "polite" level. 


(3) Interpersonal Diagnostic Report. This patient presents a pas- 
sive, dependent picture at Level I-M. He is much more depressed, 
worried, and anxious than the two preceding patients; that is to say, 
he is much more neurotic at the symptomatic level. 

His conscious self-perceptions center around weakness, docility, 
and agreeability. He tends to be hard on himself. He completely de- 
nies any strong, generous feelings (which were claimed by the two 
preceding "sicker" cases) . 

A two-layer fa9ade of extreme passivity and docile helplessness is 

The underlying tests emphasize weakness, guilt, and feelings of re- 
bellious bitterness. 

The over-all personality structure involves three layers of weak- 
ness. There is much less conflict than in the two preceding cases. 
This patient is much more like the chronic severe neurotic seen (inter- 
minably) in the outpatient clinic. He is definitely different from the 
rest of the group. He is the only patient of the six who diagnoses him- 
self (at Level II-C) as weak and needing help. He is the only mem- 
ber who stresses neurotic symptoms — depression and anxiety. When 
we recall the therapist's statement that this patient was in a "better 
state of mental health than any other group member," it becomes 
clear that internalization and expression of anxiety is a salutary security 
operation — a protection against psychosis. 

(4) Clinical Implications, (a) Motivation. This patient is well- 
motivated for psychotherapy. He experiences symptomatic pressure. 
He is dissatisfied with his personality (c.f. the self versus ideal dis- 

(b) Prognosis. A rigid and deep-seated commitment to passive, 
masochistic security operations (665) suggests a slow prognosis. This 
patient tends to avoid (at all levels) strong, responsible behavior. 
Negative identifications and underlying guilt will make therapy a very 
long-term proposition. 

(c) Predicted interpersonal behavior. The top level scores (66) 
predict a docile, conforming, placating fagade and an avoidance of 
hostile relations. They correlate with the clinician's impression of a 
"nice guy." The tests ignominiously fail to predict his going AWOL. 
We should expect masochistic self-effacement. The TAT does pick 
up some rebelliousness but does not forecast an active disaffiliation. 

There is an interesting side issue which develops from this case 
history. This patient is the only member of the group who was not 
diagnosed as schizophrenic. He was given the label severe reactive 
depression. This patient manifests the "weakest" multilevel pattern. 


He is the only patient in the group who is dependent or masochistic 
at all three levels (665), the only patient who does not claim or ex- 
press strength at some level. 

This patient was hospitaUzed not for psychotic symptoms, but for 
suicidal depression. No delusional or paranoid material was elicited. 

This patient does not fit the multilevel pattern of the paranoid 
group of five patients who claim normality or self-confidence and 
repress hostility. He stands as the representative of a second multi- 
level type which is often seen in psychiatric hospitals — patients who 
are crippled and incapacitated by a solid three-layer structure of 
despair, helplessness, and masochism. 


( 1 ) Clinical Data. This 42 -year-old patient was born in Texas, the 
youngest of fourteen children. He has resided in California since 
1937, and his present hospitalization began in 1953. The patient was 
previously committed to this hospital in 1940, when he was diagnosed 
as schizophrenic reaction, simple type. He was discharged early in 
1945 as "recovered," but was recommitted in 1953 upon the petition 
of his mother and sibhngs as: "confused . . . mumbles to himself 
. . . frightens neighbors . . . loud screaming . . . abusive." The 
patient was brought to the hospital with a black eye by pohce, who 
said that he had incurred it while resisting arrest. The patient has had 
many altercations with the law, usually occurring on occasions when 
he was intoxicated and/or driving. The patient has never married 
and disclaims any close attachments. In appearance he is lanky, dour, 
undernourished. Throughout his hospitalization the patient received 
no somatic or psychotherapy until becoming a member of the therapy 
group. In May, 1954, the patient was granted a town pass to get a 
job with leave-to-self recommended when he found employment. 
The patient would leave the grounds early in the morning and come 
back to the hospital to sleep. This continued until July, 1954, when 
the patient returned to the hospital in an intoxicated condition and 
abusive manner. He was transferred to a closed unit, where he cur- 
rently receives electro-shock therapy. 

(2) Group Behavior. The patient attended all the group therapy 
sessions and maintained a forthright and consistent position through- 
out their course, to wit, that he should not have been put in the hos- 
pital, that there was nothing whatever the matter with him, that hos- 
pitalization was more suitable to the needs of those members of his 
family who had him committed. The patient frequently stated that if 


the state wanted to support him and provide him with an easy life he 
had no objections to such a program. He was an active member of 
the group, but formed no close attachments within it. He often 
scoffed at the remarks of the other group members, and on several 
occasions shocked the other patients by making casual and crude ob- 
servations on sexual topics. His personality defenses seemed well- 
organized at all times, and he was able to maintain his equilibrium 
when pressured by other patients as to why he found it necessary to 
be such a "tough guy" at all times. The patient responded to this by 
saying that if feelings were put out in the open "somebody would 
stomp on them." This type of stomping was frequently demonstrated 
by him on the feelings of other group members. At the termination 
of the group self-leave was recommended, since he was functioning 
on a nonpsychotic level. 

(3) Interpersonal Diagnostic Report. This patient at Level I-M 
manifests a strong, unworried front. There is no attempt to present 
as a sick person — neurotic symptoms are denied. His interpersonal 
diagnosis at this level is — autocratic personality. 

The underlying tests reveal intense feelings of weakness, helpless- 
ness, and dependence. 

A fierce conflict exists between overt toughness and "preconscious" 
passivity. He cannot tolerate awareness of his" underlying fear and 
impotence. He attempts to maintain strong counterphobic operations. 

This multilevel pattern (126) seems to fit the clinical picture. He 
was able to express in group therapy his need to be strong and his 
anxieties about being seen as weak. He apparently was willing to stay 
in a protected, dependent situation in the hospital (thus satisfying his 
underlying passivity) , while stoutly maintaining the verbal picture of 

(4) Clinical Implications, (a) Motivation. This patient is not a 
candidate for psychotherapy. His counterphobic operations would be 
threatened by the implications of treatment. He has no conscious 
feehngs of anxiety or depression. He is satisfied with his adjustment 
(no discrepancy between Level II-C and Level V-C). 

(b) Prognosis. The therapeutic outlook for the 126 personality 
type is guarded. The conflict is intense, the fagade is brittle (the un- 
derlying passivity does not leak through in the form of symptoms and 
is completely avoided at Level II-C). There is the additional factor 
that underlying passivity often leads to a poor prognosis. The patient 
is warding off feelings of impotence and, in many cases, feminine 
identification. Unless ego strength is pronounced (which is not the 


case with this patient), recognition of the underlying feelings will be 
accompanied by intense anxiety. 

(c) Predicted interpersonal behavior. The test pattern (12) pre- 
dicts fairly well the therapist's descriptive title of "cynic and tough 


(1) Clinical Data. This patient is a 35-year-old divorced white 
male who was committed from a county hospital. For about two 
weeks prior to commitment the patient complained of a dust which 
kept falling from the ceiling and choking him. He was told this was 
"psycho dust" by voices which he was unable to identify, except that 
he thought doctors might be attempting to treat him from "long dis- 
tance." He was well oriented to time, place, and person. He dis- 
played no gross personality disorganization. For this reason electro- 
shock therapy was not selected as a proper treatment method, and 
psychotherapy was recommended. The patient was transferred to 
an open ward and became a member of the therapy group. 

The patient was born in Alabama and resided in the southern re- 
gion of the United States until he entered the army during World 
War II. Since his youth the patient has been a heavy drinker and was 
frequently arrested when intoxicated. On several occasions he ex- 
perienced delirium tremens and was admitted to an Alabama state 
hospital until they subsided. He performed quite well in the armed 
forces and was a technical sergeant at the time of his discharge. Since 
leaving the service, the patient has adjusted on quite a marginal basis, 
shifting from job to job and town to town, drinking heavily and work- 
ing as a seasonal unskilled laborer. 

Following group therapy the patient was being considered for self- 
leave and eventual discharge when he was transferred to a Veterans 
Administration hospital. 

( 2 ) Group Behavior. The patient attended all group sessions, and 
although he never presented any problems of his own for group dis- 
cussion, he was a very active participant. He speaks in a rather drawl- 
ing Southern accent with a dry humor which was especially effective 
because of the apt sense of timing that he displayed. He was liked 
and respected by the other group members, who referred to him as 
"intelligent." He never put forth any facts about himself or any prob- 
lems except for some very superficial facts regarding past education, 
vocational background, etc. His chief role in the group setting was 



to act as a sort of commentator, interpreting the remarks of the others 
in a humorous and sometimes penetrating manner, reminiscent of Will 
Rogers. In private conversations with the psychotherapist the patient 
expressed some anxiety about being deported to a state hospital in 
Alabama where his family lived, but he did not discuss this or any 
other area of insecurity in the group. He offered general comments 
and advice to other patients in the group, but discouraged any in- 
quiries that they might have regarding his own feelings. 

(3) Interpersonal Diagnostic Report. This patient presents a very 
mixed picture at Level I-M. He feels depressed, worried, isolated, and 
alienated. He thus internalizes his problems and recognizes emotional 
symptomology. On the other hand, he tends to emphasize physical 
symptoms and bland activity. There is, moreover, some tendency to 
maintain a conventional denial of psychopathology. An intense am- 
bivalence at this level is apparent. The tendencies to minimize emo- 
tions and to maintain strength are stronger than the admission of 
weakness. He is diagnosed at Level I-M as a narcissistic personality. 

In his self-perception he presents himself as normal and self-confi- 
dent. His diagnosis at Level II-C is competitive personality. 

His underlying tests are also conflicted, expressing strength and 

The indices from all five interpersonal tests administered to this 
patient fall in a narrow sector of the diagnostic grid — reflecting self- 
confidence and strength. There is some emotional symptomology, but 
this is minimized by the effort (apparent at Levels II and III) to act 
as a conventional, executive person. 

(4) Clinical Implications, (a) Motivation. His overt security op- 
erations tend to make him an unmotivated patient. He sees himself as 
close to his ego ideal. He admits to some emotional symptomology 
but the repressive externalizing tendencies are considerably stronger. 
He denies wanting or needing help. 

(b) Prognosis. This patient exerts a rigid control over his inter- 
personal behavior. Although he is riddled with symptoms and rele- 
gated by society to a psychiatric hospital, he still maintains a four- 
level structure of strength and conventional leadership. This rigidity 
may give him a certain stability and make it possible for him to 
function more adequately than the other group members — but the 
inflexibihty means that any major change is not to be expected. 

(c) Predicted interpersonal behavior. The solid multilevel com- 
mitment to self-confident, managerial operations (228) predicts ac- 
curately his role in the group. 



(1) Clinical Data. This parient is a 41-year-old white male who 
was born in Nevada but lived most of his life in California, where he 
obtained a university degree in business administration. He then en- 
tered the employ of a major oil company and traveled extensively in 
its service. In 1941 while on foreign duty he attempted suicide and 
was returned to the United States but not hospitalized. In 1951 he 
was briefly hospitalized because of a schizophrenic episode. He re- 
sponded rapidly to electro-shock therapy and insulin and was dis- 

At the time of his commitment the patient was married and the 
father of two children. He was extremely bitter about his hospitaliza- 
tion and very grandiose in his rationalizations. He was persecuted, 
drugged, perhaps poisoned, spied upon; the victim of greedy relatives 
and incompetent doctors. He made numerous threats and promises 
of revenge and retribution. Electro-shock therapy had no noticeable 
effect and was discontinued. The patient became a member of the 
therapy group when it was organized, but refused to continue after 
the second week. Shortly afterwards electro-shock therapy was re- 
sumed, and this time he showed improvement. He was transferred 
to the hospital annex, where his recovery progressed rapidly. He was 
described as a willing, cheerful, cooperative worker at the time of his 
release on indefinite leave of absence. 

( 2 ) Group Behavior. The patient came to the first meeting of the 
group and was by far its most active member. He assumed leadership 
of the group, questioned other patients, and steered the topics of con- 
versation to world affairs such as A-bomb strategy, etc. The other pa- 
tients in the group offered no overt objections to his taking charge, 
but the patient felt that the therapist should have asserted himself 
more than he did. During the second meeting of the group the pa- 
tient announced that he was "resigning" from it. He gave as his 
reasons the incompetence of the therapist, the poor quality of the 
hospital staff, the fact that he had been receiving poisoned cigarettes, 
etc. The therapist encouraged him to remain in the group, but could 
only elicit from him a promise to defer his decision until the next 
week. The patient never came to any subsequent meeting, although he 
was told that he was welcome to do so. 

(3) Interpersonal Diagnostic Report. This patient presents him- 
self at Level I-M as a forceful, executive, active person completely 
free from any psychological symptoms. He does not want to be seen 


as sick or isolated but, on the contrary, stresses his mental health and 
his conventional success. His diagnosis at this level is autocratic per- 

His self-perception duplicates almost exactly his symptomatic im- 
pact — although there is more emphasis on independence. His diag- 
nosis at Level II-C is narcissistic personality. 

His "preconscious" themes are loaded with superiority, rage, and 
bitterness. He is diagnosed at Level III-T as a sadistic personality. 

(4) Clinical Implications, (a) Motivation, This patient is com- 
pletely unmotivated for psychotherapy. He cannot stand any close or 
dependent relationships and maintains strong, defiant behavior at all 
levels. He is, of course, very self-satisfied and has no apparent desire 
to change himself. 

(b) Prognosis. The test pattern predicts that he will not change. 
This disagrees with the clinical history which describes his cheerful 
cooperative recovery. This discrepancy may be a test miss. There 
is some possibility that his later conforming behavior is a deliberate 
repressive maneuver to obtain discharge. 

(c) Predicted interpersonal behavior. The multilevel diagnosis 
(123) perfectly predicts the sequence of his behavior in the group — 
bossiness followed by an angry departure. It does not predict his re- 
covery — but there is some possibility that he has temporarily and de- 
liberately changed his tactics, not his personality structure. 


Interpersonal Diagnosis in Medical 
Practice: Psychosomatic Personality Types 

The interpersonal system of diagnosis has been developed and vali- 
dated by an outpatient psychiatric clinic. In the last chapter we have 
reported some applications of the diagnostic systems in an inpatient 
setting — the psychiatric hospital. The locale of investigation now 
moves again — this time to the office of the medical practitioner. Are 
there typical multilevel personality patterns characteristic of the dif- 
ferent psychosomatic conditions? If so, what are the functional im- 
plications of these personaUty factors? What do they mean to the 
internist who deals with these patients for treatment? 

Several intensive empirical studies have been made which provide 
tentative answers to these questions. Several hundred psychosomatic 
patients have been diagnosed by the interpersonal system. The re- 
sults indicate that the psychosomatic conditions investigated have 
typical personality correlates. The psychosomatic symptom groups 
with which we have been mainly concerned are: duodenal ulcer, es- 
sential hypertension, obesity, and dermatitis conditions (of unknown 
physical etiology). Samples of medical controls have been collected 
to compare with the psychosomatic groups. We have also followed 
the practice of comparing the behavior of psychosomatic patients with 
neurotic and psychotic patients and a psychiatric clinic admission sam- 
ple. These studies have two purposes: (1) to throw light on the 
factors which differentiate among the four major symptom types: nor- 
mals, psychosomatics, neurotics, and psychotics; (2) to test the va- 
lidity of the interpersonal system, i.e., to see if the system differen- 
tiates these groups at the several levels of personality. 

This chapter reviews the psychosomatic research which has been 
executed by the Kaiser Foundation project. At this point the results 
are far from definitive. They are being presented here not to prove 



anything about the psychosomatic groups or to claim that psychoso- 
matic patients can be diagnosed by means of personality tests. The 
study of any personality or symptom type is an enormously com- 
plex task involving multilevel patterns on large samples. In the case of 
psychosomatic groups, the external criteria themselves present taxing 
medical diagnostic problems. 

The following studies are, therefore, an attempt to illustrate the 
interpersonal diagnostic system in action on research questions. The 
multilevel analysis clarifies certain issues and raises new hypotheses. 

This chapter is outlined as follows: First, the ten samples are de- 
scribed; then, the behavior of each symptom group at Level I-M is 
presented and the results discussed; then, the same ten groups are 
compared at Level II-C. This is followed by the results at Level III-T. 
With the multilevel pattern of each group in hand, it will then be 
possible to present the typical personality structure and the nuclear 
conflicts of each of the ten important symptom groups. 

The implications for medical and psychiatric handHng of these 
cases will be included in these discussions. 


mon of Samples 


ten symptom groups to be described in this chapter are: 






1. Duodenal ulcer 



2. Essential hypertension 



3. Obesity 



4. Overtly neurotic dermatitis 



5. Self-inflicted dermatitis 



6. Unanxious dermatitis 



7. Normals (medical controls who were not seeking 
help for emotional or psychosomatic symptoms) 



8. Neurotics (in psychotherapy) 



9. Psychotics (committed to state hospitals) 



10. Random sample of psychiatric clinic admissions 

The Duodenal Ulcer Sample^ 

There are 41 patients in the Duodenal Ulcer Sample. Of these, 32 
are males and 9 females. These patients were referred from the gastro- 
intestinal clinic by internists using these criteria: (1) positive X-ray 
diagnosis of either gastric or duodenal ulcer, (2) the absence of any 

* These N's refer to the total sample. The A7's for any level are slightly different 
because some patients failed to take tests at all three levels. 

^The author is grateful to P. Raimondi, M.D., for providing the criteria and se- 
lecting the patients for the ulcer sample. 


other explanation for the finding, and (3) typical pain-food-relief 
sequence of symptoms. 

These patients had not requested a psychiatric referral; they were 
selected on the basis of their symptoms for research investigation. 
These selection factors must be taken into account. The fact that 
these patients were not involved in a psychiatric clinic referral may 
subtly influence their test responses. In all the findings reported be- 
low, it should be remembered that the ulcer sample was tested under 
circumstances different from the psychiatric clinic samples. 

The Essential Hypertension Sample^ 

There are 49 hypertensive patients in this study. Of these, 27 are 
men and 22 women. These patients were referred from the cardio- 
vascular clinic with the diagnosis of essential hypertension based on 
elaborate criteria being used for a simultaneous study of hypertensive 
diseases. In general, they are patients below the age of forty-five, 
with blood pressure consistently in excess of 145 mm Hg systolic and 
90 mm Hg diastolic, who, by means of kidney function tests, ephine- 
phrin neutralization tests, etc., were found to have no discernible cause 
for their elevated blood pressure. 

The hypertensive patients were selected for a research study and 
were not self -referred for psychiatric evaluation. These selective fac- 
tors may have influenced their attitude toward testing and therefore 
the findings should be interpreted with this possibiUty in mind. 

The Obesity Sample'^ 

The 98 female subjects who comprised the Obesity Sample were 
part of a large-scale study of obesity. The tests were administered 
before and after participation in discussion groups which lasted for 
about four months. These subjects were self-referred for weight re- 
duction and did not come for a psychiatric evaluation. The fact that 
they were not seen in a psychiatric setting may have influenced their 
responses. The fact that the entire sample is comprised of women is 
another serious limitation. In all other samples studied, males are 
stronger (but not to a significant degree) at the fa9ade levels than 
females. This factor should be taken into account when the data are 

^ The author is grateful to A. A. Bolomey, M.D., for defining the cnteria and se- 
lecting the patients for the hypertensive sample. 

* The MMPI and interaction data for the obesity sample were taken from the 
Herrick Hospital Research Project on obesity. This research, supported by Public 
Health funds, has studied several factors— dietary, physiological, and psychological— 
w^hich may be related to obesity. The psychological factors in the Herrick study 
have been investigated by Robert Suczek, Ph.D., whose theoretical and practical con- 
tributions to our work have been most valuable. 


The Three Dermatitis Samples^ 

The dermatitis samples comprise 161 subjects of which 67 are males 
and 94 females. These patients were taken from the private practice 
of a dermatologist and were tested in his office by a secretary who 
was trained in the necessary psychometric methods. 

The criterion used to select patients for the dermatitis sample was 
the presence of a skin symptom for which there exists no established 
physiological etiology. The specific symptomatic categories which 
made up the dermatitis sample were the following: 

Acne Hyperhidrotic eczema 

Psoriasis Alopecia areata 

Seborrheic dermatitis Urticaria 

Atopic dermatitis Acne rosacea 

Eczematous dermatitis Lupus erythematosus 

Pruritis Herpes simplex 

Otitis externa Warts 
Neurotic excoriations 

This sample was selected under circumstances somewhat different 
from any other sample. Subtle additudinal factors may have partially 
determined the results; consequently, the findings should be con- 
sidered with this caution in mind. 

Examination of the multilevel patterns of the dermatitis patients 
revealed that considerable differences exist among the different symp- 
tomatic groups. For example, the acne sample presents differently at 
all levels from the pruritis sample. It seems clear that skin symptoms 
do not manifest one personality syndrome. The skin is, of course, 
mediated by a complex set of physiological systems and is, in addi- 
tion, vulnerable to more external stimuli (e.g., self-inflicted excoria- 
tions) than any other organ system. 

For these reasons it seemed advisable to divide the dermatitis sample 
into three relatively homogeneous subgroups. These are tentatively 
labeled the Overtly Neurotic Dermatitis Sample, the Self-inflicted 
Dermatitis Sample, and the Unanxious Dermatitis Sample, 

The symptomatic subgroups which comprise the Overtly Neurotic 
Dermatitis Sample are acne, psoriasis, and seborrheic dermatitis. The 
psychological characteristics defining this group are: anxiety and de- 
pression. The physiological criteria are not well defined, but there 

' The dermatitis and neurodermatitis studies reported in this chapter are part of a 
large-scale study of emotional factors in dermatologic patients being conducted by 
Herbert Lawrence, M.D., Edward Weinshel, M.D., and the author. The criteria for 
defining these conditions were supplied and the selection of cases was accomplished by 
Dr. Lawrence. 


seems to be a greater involvement of the sweat or oil apparatus. There 
are 3 1 patients in this subsample: 20 females and 1 1 males. 

The Self-inflicted Dermatitis Sample includes the following symp- 
tomatic subgroups: atopic dermatitis, eczematous dermatitis, pruritis, 
otitis externa, and neurotic excoriations. This group is characterized 
psychologically by less anxiety and depression and considerably more 
underlying sado-masochism than the Overtly Neurotic Sample. The 
dermatological criteria which define this group are vague but would 
include itching, scratching, and more self-inflicted damage to the skin. 
This subsample includes 57 patients: 33 females and 24 males. 

The Unanxious Dermatitis Sample contains the following symptom 
groups: hyperhydrotic eczema, alopecia areata, urticaria, acne rosacea, 
lupus erythematous, herpes simplex, and warts. The subsample is 
characterized (psychologically) by a hypernormal facade with under- 
lying sado-masochistic trends. The physiological criteria defining this 
group are quite vague but would include circulatory and virus factors. 
This sample contains 73 subjects: 41 females and 32 males. 

The following code designations have been assigned to the derm- 
atitis samples: 

ND = 3 1 Overtly neurotic skin patients 
SID = 57 Self-inflicted skin patients 
UD = 73 Unanxious skin patients 

The scores for each of these subgroups will be presented in the 
dermatitis section of this chapter. 

The Normal Control Sample 

A group of 38 subjects, 21 male and 17 female, made up the medi- 
cal control sample. These subjects were patients seen in the derma- 
tologist's office for skin lesions for which a definite physiological 
(nonpsychosomatic) cause existed — industrial dermatitis, infections, 
skin carcinomas, etc. These patients were tested in the same manner 
as the dermatitis sample and were used as a direct control. This group 
stands as the only sample for which there is no apparent psychiatric 
or psychosomatic involvement and is, therefore, designated medical 
or "normal" control. 

The Neurotic Sample 

A group of 67 patients, 23 male and 44 female, who had entered 
and remained in psychotherapy at an outpatient psychiatric clinic 
comprise the Neurotic Sample. Of all the patients seen for intake 
valuation at the Kaiser Foundation Clinic, less than 40 per cent go 
into treatment. These tend to be patients who recognize and ac- 


cept the need for treatment. They tend to manifest openly the symp- 
toms of anxiety, fear, depression, isolation, etc. They present as neu- 
rotics and do not deny emotional symptoms as do the psychosomatic 
samples. For this reason, they have been labeled the neurotic or ther- 
apy sample. 

Motivational factors may have influenced their test results. It is 
likely that many of these patients were strongly desirous of therapy. 
They may have slanted their test responses in the direction of ad- 
mitting a greater number of neurotic symptoms. The findings should 
be studied with these factors in mind. 

The Psychotic Sample 

The Psychotic Sample is composed of patients who were com- 
mitted to a psychiatric hospital for inpatient custody and treatment. 
Six of these subjects (male) were studied while in group therapy at 
a state hospital. This group was combined with a sample of 22 pa- 
tients, 4 male and 18 female, who were evaluated in an outpatient 
clinic — diagnosed as psychotic and hospitalized. A third sample of 
20 patients from a private hospital was also included. The total psy- 
chotic sample is, therefore, comprised of 48 patients, 20 male and 28 

This sample is a heterogeneous mixture of cases. No claim is made 
that they are representative of psychotics in general. The sample is 
composed of at least two different types of psychotic patients. More 
than half are paranoid, i.e., underlying sadism or distrust covered by 
a fagade of pious hypernormality. The other group includes suicidal 
or depressed patients who have a double- or triple-level structure of 
despair, resentment, and/or withdrawal. The statistics of the Psy- 
chotic Sample combine the results from both these dissimilar groups. 
This unquestionably blurs the results. 

The Psychiatric Clinic Admission Sample 

A group of 207 patients, 73 male and 134 female, comprise the 
Clinic Admission Sample.^ This represents all the patients who ap- 
plied for diagnostic evaluation and were tested in the Kaiser Founda- 
tion Psychiatric CUnic over a six months' period. This sample is quite 
heterogeneous. It includes some severely disturbed patients, some self- 
referred persons seeking psychotherapy; but the largest majority of 
patients in the clinic admission sample were referred by physicians and 
came under the pressure of somatic or psychosomatic symptoms or 
suffering from anxiety which was not internalized or attributed to 

^ This N does not apply to Level III. At the "preconscious" level a sample of 100 
routine clinic patients was studied. 



their emorional functioning. These patients are not motivated for 
therapy and this is reflected in their test responses. 

The norms for the interpersonal diagnostic system are based on 
larger samples of clinic admissions. Thus, it is to be expected that the 
admission samples to be studied in this chapter will fall close to the 
center of the diagnostic circle (i.e., the mean). These selective and 
normative factors are important in considering the results to follow. 

Behavior of the Ten Samples and the Three Dermatitis 
Subsamples at Level l-M 

The average scores for each symptomatic group on the horizontal 
and vertical indices were obtained and plotted on a master diagnostic 
grid. Figure 33 presents the mean diagnostic placement for each 
of the eight samples at Level I-M. 

This diagram indicates that five of the groups fall in the extreme 
perimeter of the upper right-hand quadrant, thus expressing in their 
symptomatic behavior strength and conventional normality. These in- 
clude the ulcer (U), hypertensive (H), obesity (O), the unanxious 
dermatitis (UD), and normal control (C) samples. The self-inflicted 
dermatitis sample (SID) falls in the same quadrant but expresses 
slightly more passivity and weakness. 

The clinic admission sample (A) and the psychotic sample (P) 
fall close to the center of the circle. This is because both of these 
groups are composed of two types of people — those who are denying 
symptoms and stressing normality and those who are admitting weak- 
ness. The overtly neurotic dermatitis sample (ND) locates in the 
same area. 

The neurotic group (N) manifests an extreme amount of guilt and 

Table 35 presents the statistical tests which indicate the significance 
of these differences among the symptomatic groups. 

The results presented in Table 35 have considerable interest. They 
indicate that the neurotic group who openly accept and express anx- 
iety are significantly different (statistically) from every other group 
at Level I-M. 

No distinction can be made at this level between the ulcer, hyper- 
tensive, obese, unanxious dermatitis, self-inflicted dermatitis, and 
normal control samples. This means that these six groups tend to 
present the same symptomless, unanxious fa9ade and cannot be dif- 
ferentially diagnosed at this level. 

The overtly neurotic dermatitis sample (ND) is significantly more 
depressed and alienated than the six hypernormal samples. The neu- 
rotic dermatitis group is stronger and more conventional on the aver- 





^ (Hi) 

Figure 33. Behavior of Ten Samples at Level I-M. 

Code: U = Ulcer 

H = Hypertensive 

O = Obese 

C = Dermatitis Control 

A = Psychiatric Clinic Admission 

P = Psychotic 
N = Neurotic 
ND = Overtly Neurotic Dermatitis 
SID = Self-inflicted Dermatitis 
UD = Unanxious Dermatitis 

Key: The summary placement of each symptomatic group is determined by the 
intersection of the vertical and horizontal indices. The indices for each sample were 
calculated by (1) determining the number of cases in the sample falling in each of the 
eight diagnostic types (at Level I-M) and (2) feeding these numbers into the formulas 
[Vertical Index = l — 5 + .7(2 -(- 8 — 4 — 6) and Horizontal Index = 7 — 3 + -7(6 + 
8 — 4 — 2), where 1 = the number of subjects falling in the ^P-autocratic sector of 
the circle at this level of personality, etc.] The resulting indices express the central 
trend of each sample in comparison with the other seven samples. These group 
indices are not used m statistical tests. 



The Significance of Differences Among Ten Symptomatic Groups at Level I-M 





















































































NS - 

Key: These significance tests are based on chi-squares— the separation being be- 
tween the number of patients falling (at this level) in the strong conventional sectors 
(2,178) and the number of patients falling in the weak-hostile half of the circle (3,456). 
This is a crude over-aU measure which fails to pick up specific octant differences. 
More detailed splits (i.e., between pairs of octants) would increase the significance 

age than the neurotic group. It does not differ from the psychotic and 
clinic admission samples. 

These similarities and differences allow us to combine the groups 
(at Level I-M) into four categories: 

,^ J , , f Ulcer, hypersensitive, obesity, nor- 

Very strong and hypernormal |^^j ^^J/^^^ unanxious dermatitis 

Fairly strong Self-inflicted dermatitis 

Fairly weak, somewhat f Psychiatric admissions, psychotics, 

depressed and dependent Ineurotic dermatitis 

Very weak and dependent Patients in psychotherapy 

Behavior of the Ten Samples at Level II-C 

The mean Level II-C indices of the ten symptomatic groups were 
plotted on the diagnostic grid. These results are diagramed in Figure 

We observe that these results are somewhat different from the 
Level I scores. The neurotic sample (N) again falls in the passive, 
weak sector of the diagnostic circle. The chnic admission sample (A) 
and the overtly neurotic dermatitis sample (ND) again fall near the 
center. The self-inflicted dermatitis sample again falls in the upper 
right-hand quadrant indicating a moderate claiming of strong, re- 
sponsible behavior. 















V. . 

■ ^ >4sio 

^ , . l?l 1 . . 

, ■ 'r M , 


1 1 1 1 M 1 . 1 


;\ -^^^ 






FiGxniE 34. Behavior of Ten Samples at Level II-C. Code and Key: Same as for 
Figure 33. 

The normal control group (C) shifts. At Level I they manifest 
the strongest and most symptomless fa9ade. In their conscious self- 
descriptions, they do not emphasize strength, they admit to some 
hostile and weak behavior and almost duplicate the sm sample. 

The ulcer group (U) also shifts. In their symptomatic presenta- 
tion (Level I-M), they stress strength and conventionality. They 
consciously claim to be tougher and more aggressive. They are (at 
Level II-C) the most independent and hardboiled of any symptom 

The hypertensives (H) stay (at Level II-C) in the responsible, 
hypernormal octant, although there is a greater emphasis on their 

The obesity group (O) remains power-oriented and produces a 
double-level fa9ade of executive strength. The unanxious dermatitis 


(UD) is the third most hypernormal managerial group at the level of 

The psychotic group (P) shows a decided shift. At Level I-M they 
are scattered between passivity and hypernormality. At the level of 
conscious self-description they become overwhelmingly hypernormal. 
This severely disturbed group thus diagnoses itself as responsible and 

Table 36 presents the statistical tests which indicate the significance 
of these differences. These results are worth comment. The ulcer 


The Significance of Differences Among Ten Symptomatic Groups at the 
Level of Conscious Self-Description (Level II-C) 





















































































Key: These significance tests are based on chi-squares— the separation being be- 
tween the number of patients falling (at this level) in the strong conventional sectors 
(2,178) and the number of patients failing in the weak-hostile half of the circle (3,456). 
This is a crude over-all measure which fails to pick up specific octant differences. 
More detailed splits (i.e., between pairs of octants) would increase the significance 

sample is extremely different from the obese and hypertensive samples. 
The latter claim more conventional hypernormal traits. The ulcer 
sample is also different from the neurotics who admit to considerably 
more passivity. 

The obesity and hypertensive groups are very similar to each 
other and are significantly different from every other sample. No 
other group approaches their strength. 

The neurotic sample is unique — no other group approaches the 
weakness they manifest. 

The panic-psychotic sample is significantly more "normal" in its 
self-diagnosis than the neurotics — and shows no difference from the 
normal dermatitis and ulcer samples. 

The similarities and differences allow us to combine the groups 
at Level II-C into the following general categories: 


,. , , , /Hypertensive, obesity, unanxious 

Very strong and hypernormal ijer^atitis 

, , J , J , , /Normals, psychotics, self-inflicted 

Moderately strong and hypernormal i, • • ^ ^ 

Very strong and aggressive Ulcer 

Very weak Neurotic 

No commitment to any modal fClinic admission, overtly neurotic 

security operation Idermatitis 

Behavior of the Ten Samples at Level III-T 

The average scores (at Level III-T Hero) for each sample were 
plotted on the diagnostic grid. Figure 35 indicates that dramatic 



^ (HI) 

Figure 35. Behavior of Ten Samples at Level III-T (Hero). Code and Key: Same as 
for Figure 33. 


shifts in behavior occur when this underlying level is brought into 

The sample, which at the underlying level manifests the most 
strength, is the normal control (C). They fall in the competitive, 
narcissistic sector of the circle. The obesity sample (O) also locates 
in this sector, but they are not as power-oriented as the controls. At 
the level of conscious self-description, it will be recalled that the 
obese sample claimed to be much stronger than the controls. While 
they do not maintain this dominance in relation to the controls, they 
remain in "preconscious" behavior the second most confident and 
independent of all the samples. 

The unanxious dermatitis sample (UD) expresses underlying 
themes of strength and hostility. This group is significantly weaker 
than the controls and significantly more hostile than psychiatric 
clinic (A) sample. The self-inflicted dermatitis sample expresses more 
underlying sadism than any other group. 

The hypertensive (H) sample is the next most hostile. This symp- 
tom group is more committed to underlying sadistic feelings than t